19 February 2013

Calories aren't right on labels and maybe that’s OK

Has it ever crossed your mind that the number of listed Calories (Kcals) of, say, a large, raw, whole apple at 116 Kcals and that of a glazed doughnut at 125 Kcals might not be an accurate comparison*? Surely, you might think, isn’t the doughnut more likely to add inches to your waistline?

You'd be right. The difference that you might have understood intuitively is that, although the number of listed Kcals are similar, your body is likely to extract more of them from the doughnut than the apple. Why the disparity then on the Kcals listing? You can lay blame on the shortcomings of the Atwater Specific-factor System.

Here's a little nutrition science history lesson: In the early 20th century, American chemist William Olin Atwater pioneered calculation of energy values from measures of heat combustion of proteins, fats and carbs. This is how we arrived to the familiar protein at 4 Kcals per gram, lipids at 9 Kcals per gram, and carbs 4 Kcals per gram. It would come to be known as the Atwater General Factor System. At the time, Atwater couldn’t account for fiber, so, in 1955, Bernice Watt and Annabel Merrill refined the system with specific Calorie conversion factors of foods, which has led to what the system is now.

But the Atwater Specific-factor System still doesn't take a lot of other variables into account. For example, it doesn’t consider digestion energy costs of one food versus another, such as those from chewing, secretion of gastric acid and digestive enzymes, intestinal movement (peristalsis) and production of heat after eating (diet-induced thermogenesis). It also doesn't account for losses to your friendly gut microbes, which can devour up to half of calories from any digestion-resistant starches (as found in a raw apple) that arrive in the large intestine after escaping breakdown by gastric acid and pancreatic enzymes in the stomach and small intestine.

Now, in the wake of news that McDonald's and other restaurants will start listing the numbers of calories next to menu items—a requirement under a new regulation included under the Affordable Care Act—comes a fresh challenge on the Atwater Specific-factor System and those Kcals listings for being inaccurate. That challenge is from an unlikely group of experts from diverse fields that include evolutionary biology and comparative physiology on mice and pythons. (Yes, pythons! I should spare you the yucky details, but I won't.)

On Monday, February 18, at the American Association for the Advancement of Science annual meeting (Twitter hashtag: #AAASmtg) in Boston, the group convened to summarize their findings. In addition, the group announced, they hope to write a position paper on the topic within the next few months. In some cases, the group reports, Kcal counts on food labels could be off as much as 15 percent or more.

Evolution and pythons

Richard Wrangham and Rachel Carmody of Harvard are not part of the so-called "nutrition establishment"; they are a pair of biological anthropologists who study primates and human evolution. Over the last decade and a half, Wrangham has studied the role of fire in the evolution of humans (as regular readers of my blog surely are aware). Carmody's research is in evaluating what effect basic food processing (e.g. pounding) and cooking (as adopted by our ancestors probably beginning with Homo erectus) had on freeing up available Kcals from raw foods so as to drive increase in survival and human adaptations including larger brain size.

Carmody detailed what was the first Harvard lab experiment on effects on energy gain that mice experienced when they ate foods—sweet potatoes and lean beef—that were either raw and whole, raw and pounded, cooked, or pounded and cooked. She found that pounding increased energy gain of the meat by 8 percent, cooking by 15 percent. Pounding increased energy gain from the tuber by 3 percent, cooking by 39 percent, the combination by 40 percent. The mice also showed greater preference for foods that were pounded and cooked after the trial.

In addition, Carmody presented preliminary data that cooked versus raw food made an impact on gut microbial communities and genetic expression related to nutrient metabolism and immunity—the raw beef induced the most expression from immunity-related genes (possibly in response to pathogen load), which costs more energy. As hypothesized by Carmody, the pounding and cooking reduced diet-induced thermogenesis in the mice and, thus, reduced energy expenditure.

Diet-induced thermogenesis, aka specific dynamic action (SDA) as it's referred to often by comparative physiologists, is the subject of research by biologist Stephen Secor of University of Alabama. "There's no such thing as a free meal," Secor likes to remind. "Every time you consume a meal, there's a metabolic expense." Again, that is not accounted for in the Atwater system.

Secor, whose research is in reptiles, explains that contributions to SDA involve pre-absorptive (chewing, swallowing, bile and pantreatic secretions, peristalsis) and post-absorptive (nutrient breakdown and transport) phases. The gastric acid production, however, can heavily influence digestion costs—up to 25 percent for some foods—and varies greatly depending on the animal. "There are more mitochondria in cells responsible for acid production than anywhere else," Secor said.

X-ray image of a Burmese python while digesting its meal.
OK, about the pythons: when pythons eat a rat whole, Secor says, they have a huge gastric acid and SDA response (they get so hot, it's like "they're on fire"). That makes them a perfect model to study the effects of processing and cooking as it relates to reduction in SDA. To study these effects, Secor's lab wrapped a steak around a rat and fed it to the snake (apparently, "you can feed anything to a python as long as you put a rat's head on it"). The research group also blended up a rat smoothie and delivered it via tube down a snake's throat and deposited it directly into the snake's stomach (bypassing the swallowing phase). From these experiments, Secor estimates that cooking and grinding reduces SDA by a whopping 25 percent.

Knowing what we do about how the size of meals increasing gastric secretion, especially those containing protein, Secor's findings suggest there are greater digestion costs on a less processed or raw diet, particularly if consisting of large meals with protein.

The fate of carbs

Cooking and processing also strongly affects the gastrointestinal fate of carbohydrate in meals and their energy contributions. When starch is provided as intact granules, it can be resistant to hydrolysis from enzymes during digestion, explains Klaus Englyst of Englyst Carbohydrates Ltd, Southampton, UK. Once heated, especially if moisture is involved, the starch gelatinizes allowing amylase (starch enzyme) to more easily break it down for rapid digestion and absorption. The resistant starches, on the other hand, might end up in the intestine where microbes will ferment it using some of its energy up and producing short-chain fatty acids (which we absorb).

Englyst hints that resistant starches and fibers (cellulose, beta-glucan, etc.) deserve more attention for their possible role in providing the protective effects as seen in studies on low-glycemic diets. In fact, he said he's concerned that the term low glycemic in itself is an inconsistent message. Low glycemic refers to a diet based on the glycemic index (as developed by David Jenkins of University of Toronto) that quantifies blood glucose level rises after eating specific foods. One could argue that a hamburger is low glycemic, since it may not spike blood sugar, yet it contains very little resistant starch or fiber.

Despite shortcomings of low glycemic as a term overall, Dr. David Ludwig, a physician at Boston Children's Hospital (who filled in on behalf of Peter Turnbaugh at the conference), says it's useful for discussing the different kinds of carbohydrates. Unprocessed grains have fiber and intact grain structures causing pancreatic enzymes to really have to work at breaking them down, whereas today’s highly processed grains have their kernels milled to fine particulate and fiber stripped away. That’s how they become high glycemic (able to spike blood sugar). "You're left with Wonder Bread and there are thousands of iterations of this product," he said.

He also shares the view that the USDA's "low-fat" message and the Food Pyramid (with grains as its base) has been harmful over the last few decades. "You can take fat out, replace it with highly processed starch molecules that are immediately susceptible to attack from amylase," Dr. Ludwig said. Last June, Dr. Ludwig and his colleagues published a study in JAMA that found that a low glycemic diet made it easier for subjects to maintain weight loss over time as compared to a low-fat or low-carb diet.

Dr. Ludwig also hinted that there may be more to the low-glycemic story in ways beyond greater digestion costs of resistant starches and fibers and into the bloodstream. Interesting new findings, he told me after his talk at the conference, suggest high-glycemic meals could have an impact on hormones in terms of how they influence appetite, metabolic rate, and energy storage (I wonder what he thinks of "insulin hypothesis").

Nutritional biochemist Geoff Livesey, of Independent Nutrition Logic Ltd., is the lead author of a meta-analysis of prospective cohort studies that Dr. Ludwig told me was the most well-designed study available on low-glycemic diets to date. The study, just published in American Journal of Clinical Nutrition, indicated a "strong and significantly lower type-2 diabetes risk" in those consuming low-glycemic diets. Livesey added that it’s unclear whether or not it’s the lack of highly processed carbs and fats that make the low-glycemic diet protective, or the inclusion of more protein and fiber that produce the protection; protein and fiber, for example, have a greater satiety effect than do carbs and fats that could contribute to an easier time eating reduced calories.

Livesey agreed with the panel that the Atwater Specific-factor System had failures and could be improved to be more accurate, but he cautions that the group had an uphill battle if they were going to change it. It would take changing the position of the Food and Agriculture Organization of the United Nations (FAO). Also, many in the "nutrition establishment" (Ludwig's choice of words) are likely to disagree that the Atwater system needs to change at all.

Does it matter if food labeling is inaccurate?

The FAO, in fact, brought together an expert committee to write a report after a technical workshop in Rome that evaluated the topic of food energy and the Atwater Specific-factor System in 2002. In the report, the committee concluded that, yes, it was true that foods can differ substantially in terms of their "net metabolizable energy." But more data need to be collected before determining any major changes and, in the context of the total diet for most countries, the Atwater Specific-factor System isn't likely to introduce a lot of error—they estimated less than 5 percent, except on a few subsistence diets, which rely on a lot of native foods, where errors may increase to up to 18.5 percent.

I asked John Peters, Chief Strategy Officer of the New Health and Wellness Center at the University of Colorado Anschutz Medical Campus, to comment further on the topic. He told me, "Every step of the digestion and metabolism process introduces multiple variables that will be unique to the individual." For example, he told me, the "digestible energy of lard (made of palmitic, stearic and oleic acids) will depend on the calcium level in the diet of the organism consuming the lard because calcium will bind to stearic acid in the gut and reduce its absorption."

Individual variability adds complexity that is compounded by the metabolic energy costs that are heavily determined by protein content of meals because of the need to remove amino acid waste, along with fermentable fiber content, rate of absorption of foods into the body, and other elements of nutritional status. "As you can imagine," he told me, "these are variable from meal to meal and from individual to individual. Because of this complicated mix of effectors, the current labeling system was an attempt to blend simplicity with the 80:20 rule… it's about right for most things found in most diets.”

Mark Haub, a professor of human nutrition at Kansas State (aka the Twinkie diet professor), explains that his belief is that people misunderstand or simply fail to remember the components of thermodynamics. "One factor is that food contains a constant amount of energy, and then the potentiallly incongruent piece is that the energy in the food may or may not be bioavailable" (as metabolizable energy). So, a processed food would, indeed, tend to have energy that is more bioavailable than raw food because the energy is more difficult to extract because it's more difficult to digest.

Yet he agrees that when it comes to changing food labeling, "that becomes tricky." The absorption of energy from oranges, for example, can change simply depending on how much is eaten. When large amounts are eaten, more will likely "get dumped" into the large intestine. In addition, Haub told me, the rate of digestion can change depending on accompanying foods like those containing fiber or protein. "Do we digest an apple the same every time we eat it? I have a hard time thinking that is the case."

He's right about the differences in apple absorption considering that people are likely to eat them raw sometimes, processed (as in apple sauce), and cooked (as in apple pie on Thanksgiving). When eaten raw, according to Carmody's findings, it would yield fewer Kcals. But since the Atwater Specific-factor System basically assumes all foods are processed or cooked, at least consumers can safely predict that the apple would not yield more than the listed 116 Kcals.

So, perhaps that's a better way to look at the problem: labels wouldn't be so misleading if we thought of their Kcal listings not as total Kcals, but as max Kcals you could get from any given food. And, if one is to determine some kind of main takeaway dietary message from all this discussion, it may be simply this: that by eating more whole, raw foods (like other animals and like our pre-human ancestors did), you can count on fewer Kcals (thank your microbes for halving the amounts of Kcals you absorb once food reaches your intestine); and you can count on greater digestive and metabolic costs by eating well-balanced meals containing plenty of fiber and protein per meal (think "python diet").

*Kcals from USDA's nutrient database.

Photo credits: iStockphoto and Inside JEB.


Puddleg said...

That's a very good discussion of the calorie issue.
Depending on your health needs you might not want to cheat yourself of nutrients by eating raw. It won't always be better, though it may help weight loss IF you're already nourished.
Barry Groves has stated, in Trick and Treat, that the true caloric value of triglycerides depends on their chain length, because glycerol has the lower value of a carbohydrate. The longer the fatty acids in a triglyceride, the lower the proportion of glycerol. Thus, say, coconut oil MCTs will have significantly fewer calories per gram than very long chain fish oil triglycerides.

Mary M said...

What a great dissection of this issue. Thanks very much. I saw the meeting tweets and was very curious about what the details were.

Jeannine said...

Excellent post! Especially liked the hot snake and McDs -- so much info to process. Thanks!

Mark Haub, PhD said...

Well done David! I hope your "max kcal" concept sticks.

Mark Haub, PhD said...

Well done David! I hope your "max kcal" concept sticks.

Brian MacNamara said...

Great article, makes perfect sense in the same way as cooking or processing a food changes its Gi value and affect on blood sugar and insulin response. Its naive to expect that the body handles different foods with different nutrient contents in the same way - the fact that there's complex biochemical processes at work in our body mean that the reactions will differ depending on what the inputs to it are. That just simple Science 101.

Mark Haub, PhD said...

Let's not forget that processing tends to make ALL nutrients (micro and macro) more bioavailable. Thus, the desire to minimize processing may negatively impact availability of other essential nutrients (iron, vitamins, etc). Here is an example:


daviddespain.secret said...

Thanks for the comments, everyone.

Dr. Haub makes an excellent point about micronutrients becoming more bioavailable. We often hear that "raw foods have more micronutrients" and that processing/cooking removes them to a degree. However, beta-carotene is an excellent example of how micronutrients can be more bioavailable with processing. There are also other examples such as folic acid (in cereals/pasta) and lycopene (in tomatoes) that become more bioavailable when processed or used to fortify processed foods.

Miki said...

I fail to understand the merits of dealing with calories one way or another as it concerns public health. It just distracts us from dealing with the real possible problems like over-consumption of gluten, fructose and pufa or the under-consumption of fiber and animal fat. With proper nutrition there is no biological/evolutionary reason why we would consume one extra calorie and with improper nutrition no amount of calorie-counting will save us.