I love nutrition science. So much so that I read a new book on it every week. The Alternative Hypothesis has a pretty old video on agriculture and evolution. I strongly disagree with his main thesis. I strongly disagree with his denigration of Gary Taubes. Most of all, I strongly disagree with what he says about the East Asian rice eaters because since that video has been made, the carbohydrate/insulin hypothesis of obesity has changed to the insulin hypothesis of obesity.
In the very beginning of the video he brings up Gary Taubes’s research on low-carb diets and how people tend to be healthier than those who eat higher carb diets. He brings up the East Asians who eat a lot of rice. However, it’s clear he doesn’t know that the percent of carbohydrate intake is nowhere near as important as the absolute amount of carbohydrate consumed:
- They consume a fraction of the sugar we do. More sugar consumption leads to greater insulin resistance, more fat creation, less fat breakdown, and more fat accumulation.
- They consume less total glucose, AND the glucose they consume is accompanied by less sugar (and less omega-6 PUFA, if it matters).
- They consume a ratio of omega-6 to omega-3 PUFA that is much lower than we do. This mayfurther reduce any insulin resistance brought on by the glucose they do consume (in smaller doses and with less sugar).
The fact that East Asians didn’t have high rates of diabesity (diabetes and obesity) was a big blow to the carbohydrate insulin hypothesis. However, the East Asian rice paradox is simply explained by low, if non-existent, consumption of refined carbohydrates. Those populations actually consume fewer total carbohydrates than Western diets, and have lower levels of glycemic load as a result. To quote Mark Sisson:
Before recently, Asians ate less refined sugar and used animal fats for cooking. Sugar intake is rising now, of course, and cooking oils made from corn and soybean have largely replaced lard and tallow, but rice in the context of a low-sugar, no-HFCS (remember, the oft-cited 55/45 fructose/glucose breakdown for HFCS is highly misleading and actually quite often incorrect), low-vegetable oil, nose-to-tail nutrient-dense diet is (or was) acceptable. You can’t reduce a food down to its constituent parts and focus on, say, the bit of fructose in a blueberry and then condemn the entire berry because of it. Similarly, you can’t reduce a diet down to a single constituent food and condemn – or praise – it based on that single food. You have to look at the entire picture, and the Asian diet is largely a nutritious one.
These paradoxes where one population seems to disprove a certain hypothesis are pretty easily explainable with the existing data. There are numerous reasons why East Asian rice eaters have lower rates of diabesity. Dr. Jason Fung also explains why:
Wheat, particularly in the modern iteration may be particularly fattening for numerous reasons. The high level of amylopectin means that most of the starch contained in flour is efficiently converted to glucose. This deadly combination of wheat and sugar has been introduced into the Chinese diet. The result is a Chinese diabetes catastrophe. The prevalence of diabetes in China has now even outstripped the USA.
This is the answer to the paradox of the Asian Rice eater puzzle. Why didn’t the Chinese have a diabetes epidemic in 1990 with all their white rice? Well, because they didn’t eat any sugar (fructose), they were not developing insulin resistance. Because they were not snacking all the time, they had periods of low insulin level that helped prevent the development of insulin resistance. So the high rice intake by itself was not enough to cause either of diabetes or obesity.
Then he says that whites intake more total carbs in comparison to blacks and ‘Hispanics’ (1:32 in the video). This is wrong.
Diaz et al (2005) showed that minority populations are more likely to be affected by diabetes mellitus which may be due to less healthy diets and/or genetic factors. Using the National Health and Nutrition Survey for 1999-2000, they analyzed overweight, healthy adults, calculating dietary intake variables and insulin sensitivity by ethnicity. They characterized insulin resistance with fasted insulin, as those who are more likely to become insulin resistant have higher fasted insulin levels (levels taken after waking, with the subject being told not to eat the night before as to get a better reading of fasted insulin levels). Non-‘Hispanic’ whites had higher energy and fat intake while ‘Hispanics’ had higher carb intake with blacks having lower fiber intake. Blacks and ‘Hispanics’ were more likely to have lower insulin sensitivity. However, ‘Hispanics’ were more likely to have lower insulin sensitivity even after controlling for diet, showing that metabolic differences exist between ethnicities that affect carbohydrate metabolism which leads to higher rates of diabetes in those populations.
Diaz et al state in the results of the study:
Dietary differences are seen by ethnicity, with non-Hispanic whites having higherenergy, saturated fat and total fat intake, while Hispanics had higher carbohydrate intake and African-Americans had lower fibre intake.Both African-Americans and Hispanics had higher levels of fasting insulin, demonstrating lower insulin sensitivity in comparison with non-Hispanic whites.
So now that I’ve established that blacks and ‘Hispanics’ consume more total carbohydrates from refined foods, now I’ll show the physiologic effects of insulin.
Insulin inhibits the breakdown of fat in the adipose tissue by inhibiting the lipase that hydrolyzes (the chemical breakdown of a compound due to a reaction with water) the fat out of the cell. Since insulin facilitates the entry of glucose into the cell, when this occurs, the glucose is synthesized into glycerol. Along with the fatty acids in the liver, they both are synthesized into triglycerides in the liver. Due to these mechanisms, insulin is directly involved with the shuttling of more fat into the adipocyte. Since insulin has this effect on fat metabolism in the body, it has a fat-sparing effect. Insulin drives most cells to prefer carbohydrates for energy. Putting it all together, insulin indirectly stimulates the accumulation of fat into the adipose tissue.
Do you see why blacks and ‘Hispanics’ are more susceptible to obesity?
Another glaring error he commits is not separating refined carb consumption with natural carb consumption. Refined carbs spike insulin much more than those foods with natural carbohydrates. East Asians do not have a “higher carbohydrate tolerance than Europeans” (2:06 in the video). This one huge error he commits completely discredits his hypothesis.
He then goes on to talk about India’s diabetes rates. But why is it increasing? Because of Western diets. It’s not about a “lower carbohydrate tolerance” as he says at 3:07, it’s about consuming more refined carbohydrates.
Then at 5:05, he says that he’s “solved Gary Taubes’s race problem in regards to diet”. He did nothing of the sort.
I, of course, have no problem with his IQ data. I have a problem with the conclusions he jumps to in regards to carbohydrates and diabetes. He clearly didn’t look at other factors that would explain why East Asians have lower rates of diabesity (which is increasing as they adopt a Western lifestyle… Weird…). The same thing explains it with the Australian Aborigines.
I have absolutely no problem with the second half of his video. My problem is the first half of it–his denigration of Taubes, non-understanding of insulin spikes in comparison to the quality of carbohydrate ingested and not controlling for refined carbs– as it’s clear he didn’t do extensive research into these populations (which Taubes and others have) to show why they don’t have higher rates of diabesity.
What he doesn’t touch on are “obesogenic environments” which is defined as “the sum of influences that the surroundings, opportunities, or conditions of life have on promoting obesity in individuals or populations”. What a huge coincidence that most of the populations he cited today with higher rates of diabesity live in first-world nations, otherwise known as obesogenic environments.
He should have spoken about the Pima Indians and their rates of diabesity. They didn’t have rates of diabesity as high 100 years ago. Why? The introduction of the obesogenic environment. Prisancho (2003) in his study on the Pima and reduced fat oxidation in first-world countries showed how the Pima preferentially burn carbs and not body fat for energy. Fat-burning would account for 9 kcal lost and CHO for 4. Since they preferentially burn carbs for energy and not fat, this shows why they have higher rates of diabesity. It’s not that it’s a genetic susceptibility to burn CHO for energy over fat (there may be a small genetic component, but it doesn’t override the effects of the actual diet). I’ve shown insulin’s role in fat storage above, do you see why the Pima have this diabesity epidemic after the introduction of refined carbohydrates and the obesogenic environment?
Added sugars and salts in foods causes us to want more of those foods. As I alluded to above, food scientists continuously work to find out which combinations of sugar, salt and fat will be more hyperpalatable to us and make us eat them more. Whites nor East Asians have a ‘higher carb tolerance’, they just eat different types of carbs (mostly unrefined, in comparison to blacks and ‘Hispanics’ anyway). If any individual were to overeat on high carb foods they would become diabetic and obese. Whites nor East Asians are exempt from that.
In sum, he didn’t look at where the carbs came from, only total carb intake. Refined carbs and unrefined carbs do different things in the body. The whiter a processed food is, the more it is refined. The more a food is processed, the more its natural nutrients such as fiber are taken out. These low-fat refined foods are one cause of obesity. However, it’s way too complicated to say that only refined carbohydrates cause diabesity.
I strongly recommend he read Taubes’s and Fung’s books. If he did, he wouldn’t have said what he said about Taubes’s theory and completely disregard the absolute total amount of carb intake and not the total amount of carbohydrates ingested.