http://www.ncbi.nlm.nih.gov/pubmed/20837645
This is a great example of how complicated it is to do good science on humans, including how assumptions can confound an otherwise interesting study.
The abstract starts off with this interesting assumption: "A fat-rich energy-dense diet is an important cause of insulin resistance." I don't have the subscription to the publication journal, but I would really love to see how they justify this statement. What's their definition of "fat rich"? What exactly are the parameters of "energy dense"? What's interesting about these terms is the implications in a real person's diet. For example, try to eat a fat rich, low carb diet, and you find you don't eat a great deal of total "energy". Why? Hunger is very well controlled on a fat rich, carb restricted (50g/day or less) diet. If you eat a small amount of an "energy dense" food like coconut oil, that WILL NOT make you insulin resistant. Ever eat a large quantity of butter without any sugar? Probably not, with the possible exception of macadamia nuts! (In fact, if you overeat fat by too large a margin, you'll have enough diarrhea to discourage this practice). I'd venture a guess that, assuming you could eat "too much" coconut oil/butter/cheese/full fat dairy, it still would not make you insulin resistant (by the way, I'm betting my life on this answer because that's how I eat). What makes a person insulin resistant is eating a bunch of grain products, whole or otherwise, and/or food laden in sugars (and fructose is specifically implicated in the genesis of insulin resistance - beautiful explanation of why is available at this link: Lustig on Fructose). The fact is that we often eat a lot of fat when we're eating a lot of grains/sugars, but that's a different case entirely than when one eats low carb (50g/day or less), adequate protein (minimum 60g/day), and good fats to satiety.
Switching back to the study, I'm skeptical when anyone starts off with terms like "energy dense" or "fat laden" because this is the language of those who have led the science of diet into demonstrably poor science for the last 30+ years, the same flavor of science that has, incredibly, accepted the conjecture that high fat diets cause heart disease, and that weight gain can be simplified to a restatement of the First Law of Thermodynamics (which is not relevant for determining causality in human weight gain).
The study took three groups, all of which were fed a "Healthy male volunteers (18-25 y) received a hyper-caloric (∼+30% kcal day(-1)) fat-rich (50% of kcal) diet for 6 weeks." In other words, they tried to over-feed the subjects by 30%, using a diet of 50% fat (which they think is a high level of fat).
In short, one group didn't exercise and gained six pounds. Another group exercised in a fasted state and gained almost no weight. The third group ingested carbs prior to exercise (IOW, exercised but not in a fasted state) and wound up about in the middle of the three groups. The three groups had similarly different outcomes in terms of other measures of health (insulin sensitivity, et al).
Interesting result - two groups had equal intake, and equal exercise but different outcomes for weight gain and health markers. Why? How could the "calorie is a calorie" model be true if this study proves repeatable (and therefore valid)?
In other words, what use can we make of the implications of this study? That's a tough question, because as is so often the case, there are many "complexifiers." For example - what would have happened had they used only protein/fat in the pre-workout "fast break"? What would have happened if they used 100g/day or less of carbohydrate? What if the long endurance training sessions had been either shorter, or shorter but more intense (anaerobic), or had incorporated resistance training? The mental experiments we could imagine would fill up a large chart with possible options to be tested. Unless you think it's a great idea to burn up four hours a week doing pointless and possibly injurious cardiovascular training, knowing that you can "overeat" by 30% if you train in the fasted state is not a useful data point.
In part 2, I'll superimpose this study result with the Paleolithic Model, to see if the two can be reconciled.
No comments:
Post a Comment