Full disclosure: I happen to love biochemistry. I have a favorite transcription factor (ChREBP) and a favorite neurotrophic factor (BDNF). I think proteins are beautiful. If I were a biochemist who had discovered a novel protein, I would carry a picture of it around with me in my wallet.
The animal and cells models used in biochemistry are great for looking at genetics, epigenetics, at biological mechanisms, and how these things interact. We can manipulate these models in ways that we can’t with humans, and this has given us some crucial insights into mechanisms, especially neural and epigenetic ones—critical to understanding the effects of nutrition—that would be virtually impossible to study in humans.
Nutritional biochemistry can also wear the mantle of “objective-er than thou” when it comes to science. As one of the biochem profs at UNC noted: If you have to use statistics to discuss the results of your experiment, you need to redesign your experiment. Sure, the questions asked, the interpretation of results, and what gets published in biochem are influenced by funding sources, social/scientific contexts and dominant paradigms. But unless you are a truly bad scientist, you can’t make the experimental results come out in a way that supports your hypothesis.
(This is in marked contrast to observational studies in nutrition epidemiology where the whole point of the data analysis “experiment” is to find results that support your hypothesis. Sometimes you don’t find them, and those findings should be reported, although they may not be because who’s to know? Just you and your SAS files. My point is that you are actively seeking results that confirm a particular idea, and this just might influence what “results” are found. More on this in another post.)
But beyond the utility and elegance of nutritional biochemistry, the problems with regard to health policy are two-fold.
The first problem: In many ways, nutrition policy has become almost completely divorced from the basic science investigations done in biochemistry. The Dietary Guidelines Advisory Committee (DGAC)—the committee of scientists that, at least theoretically, reviews the science upon which the US Dietary Guidelines are based—started in 1985 as mostly MDs and biochemistry professors. As time went on, the DGAC became more heavily populated with epidemiologists. This would be fine if epidemiology was meant to generate conclusive (or even semi-conclusive) results. It isn’t. Epidemiology gives us associations and relationships that are meant to be understood through a reasonably plausible, preferably known, biological mechanism. Note these interesting conclusions from the 2010 DGAC Report and the 2010 Dietary Guidelines policy document with regard to dietary cholesterol:
Here’s our mechanism: “Exogenous, or dietary, cholesterol down-regulates cholesterol synthesis in the liver to maintain cholesterol balance.”
[D3-1, Reference 1, emphasis mine]
Here’s our epidemiology: “Traditionally, because dietary cholesterol has been shown to raise LDL cholesterol and high intakes induce atherosclerosis in observational studies, the prevailing recommendation has been to restrict dietary cholesterol intake, including otherwise healthy foods such as eggs.”
[D3-2, Reference 1, emphasis mine, "induce"? really? how does one "observe" that cholesterol "induces" atherosclerosis? I'm assuming committee fatigue had set in at this point because that word should have been "are associated with"]
Here’s our policy recommendation: “Consume less than 300 mg per day of dietary cholesterol.”
[Ch. 3, p. 21, Reference 2]
See, wasn’t that easy?
This brings me to the second problem, which is sort of the flip-side of the first: Biochemical processes that are understood primarily through mouse or cell models only work as the basis for dietary recommendations for chronic disease if you’re making them for cells or mice.
As one of my favorite professors in the Nutrition department likes to quip, “We know how to cure obesity—in mice. We know how to cure diabetes—in mice. We have all the knowledge we need to keep our rodent population quite healthy.” Obviously this knowledge has not been translatable to humans. In some ways, basic nutrition biochemistry should be divorced from public health policy.
The reason for this is that the equivalency of animal models to humans is limited in ways that go beyond simple biological comparisons—although the biological differences are significant.
My knowledge of comparative physiology is limited at best, but my understanding is that most rodents used in nutrition biochemistry work (rats included) have a cecum (an intestinal pouch that facilitates the breakdown of cellulose), an adaptation that would be necessary in a diet composed of hard-to-digest plant material such as seeds and grains. Because this process is not terribly efficient, many rodents also recycle nutrients by eating their feces. Humans don’t have a functional cecum for fermentation; we don’t tend to reingest our own poops (or anyone else’s poop, unless you’re starring in a John Waters film) in order to extract further nutrition from them as our bodies are already very efficient at this during the first go-round.
Furthermore, due to inherent difference in physiology, animals may not accurately model the physiological conditions that produce disease in humans. For example, in some species of rodents, a high fat diet will induce insulin resistance, but there is no definitive evidence that higher fat intake per se impairs insulin sensitivity in humans . Why this is so is not entirely clear, but likely has something to do with the diet each species has consumed throughout its evolution. In a natural setting, rodents may do well on a diet of mostly grains. On the other hand, humans in a natural setting would do okay on a diet of mostly rodents.
What is more critical is that animal and cell life can’t imitate the complex environmental inputs that humans encounter throughout their lives and during each day. Animals and cells only get to consume what they are given. If you’ve ever been at a conference where the breakfast is low-fat muffins, whole grain bagels, fat-free yogurt, orange juice, and fruit, you know what that feels like. But typically our food choices are influenced by a multitude of factors. Mice, unlike humans, cannot be adversely affected by labeling information on a box of Lucky Charms.
Does that matter? You bet it does.
Where do most Americans get their nutrition information these days? From media sources including the internet, from their grocery stores, from the packages holding the food they buy. People who have never read a nutrition book, much less the actual Dietary Guidelines, still “know” fat is bad and whole grain is good [4, 5]. These environmental exposures affect food choices. Whether or not the person still decides to consume food with a high fat content depends on another set of cultural factors that might include socioeconomic status, education, race or ethnicity, age, gender—in other words, things we can’t even begin to replicate in animal or cell models.
Human biochemistry is unique and complex, as are our social and cultural conditions, making it very difficult to study how these primary contributors to health and food choices are related to each other.
Can we do a better job with nutritional epidemiology? I know you’re on the edge of your seat waiting for the next episode in the unfolding drama, N of 1 Nutrition, when we get to hear Walter Willett say:
“I never met a statistical man I didn’t like.”
1. U.S. Department of Agriculture. Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans 2010. Accessed July 15, 2010. http://www.cnpp.usda.gov/DGAs2010-DGACReport.htm
2. U.S. Department of Agriculture and U.S. Department of Health and Human Services. Dietary Guidelines for Americans, 2010. http://www.cnpp.usda.gov/DGAs2010-PolicyDocument.htm Accessed January 31, 2010
3. Report of the Panel on Macronutrients, Subcommittees on Upper Reference Levels of Nutrients and Interpretation and Uses of Dietary Reference Intakes, and the Standing Committee on the Scientific Evaluation of Dietary Reference Intakes. Dietary Reference 4. Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty Acids, Cholesterol, Protein, and Amino Acids (Macronutrients). Washington, DC: The National Academies Press; 2005.
4. Eckel RH, Kris-Etherton P, Lichtenstein AH, Wylie-Rosett J, Groom A, Stitzel KF, Yin-Piazza S. Americans’ awareness, knowledge, and behaviors regarding fats: 2006-2007. J Am Diet Assoc. 2009 Feb;109(2):288-96.
5. Marquart L, Pham AT, Lautenschlager L, Croy M, Sobal J. Beliefs about whole-grain foods by food and nutrition professionals, health club members, and special supplemental nutrition program for women, infants, and children participants/State fair attendees. J Am Diet Assoc. 2006 Nov;106(11):1856-60.