The 4-Hour Body_ An Uncommon Guide to Ra - Timothy Ferriss [202]
To the greatest extent possible, avoid studies that depend on after-the-fact self-reporting. Trust your own data. Just record it when things happen.
4. IS THIS DIET STUDY CLAIMING A CONTROL GROUP?
Desirable as it may be, it is almost impossible to change just one macronutrient variable (protein, carbohydrate, fat) in a diet study. It is therefore almost impossible to create a control group.
If a researcher makes such a claim and vilifies a single macronutrient, your skeptical spider sense should tingle.
Let’s suppose that someone claims that a study on the effects of a low-fat diet proves it’s healthier than a higher-fat diet. There is a control group.
First things first: if you decrease fat or cholesterol in a whole-food diet, you’ll be removing fat and protein, which means you must add carbohydrate to make the diets equal in calories. If you don’t, you have unequal calories as another variable. If you make this caloric correction by adding carbohydrate elsewhere, now you face a conundrum: not only are you comparing a higher-fat and a lower-fat diet, but you’re simultaneously comparing a higher-protein and a lower-protein diet, as well as a higher-carb and a lower-carb diet.
How can we know which is responsible for what?
We can’t.
Self-experimentation actually offers an unexpected advantage here.
The “control” is everything you’ve tried up to a certain point that hasn’t produced a desired effect. Isolating one variable is often less important than the sum impact of a group of changes. In other words, has your bodyfat percentage gone up or down in the last two weeks of replacing diet A with diet B? If you weren’t losing fat on A and now you are, A was your control.
In an ideal (but unattractive) test, you would go back to A and see if bodyfat then moves in the other direction. Then repeat the switch again. This would minimize the possibility that the first change in bodyfat just happened to coincide with the change in diet to B.
Alas, this switching would also maximize your likelihood of going insane. If something seems to be working, just stick with it.
5. DO THE FUNDERS OF THE STUDY HAVE A VESTED INTEREST IN A CERTAIN OUTCOME?
Beware of unholy unions between scientists and funding sources.
Fred Stare, founder and chair of the Nutrition Department at Harvard University, obtained a $1,026,000 grant from General Foods in 1960. These manufacturers of the sugar-rich Post cereals, Kool-Aid, and the Tang breakfast drink would be subsidizing the “expansion of the School’s Nutrition Research Laboratories.” In the decade that followed, Stare became the most public and reputable defender of sugar and modern food additives, all the while receiving funding from Coca-Cola and the National Soft Drinks Association, among others. Does this prove malfeasance? No. Does it show that funding, support that can be discontinued, often comes from those most likely to benefit? Yes. People respond to incentives.
Simple due diligence: Peruse the required “conflicts of interest” section in studies or reviews related to diet and look for “consulting fees.” If McCorporation or the XYZ Lobby decides it’s too obvious to fund studies directly, hiring researchers as consultants can be an indirect means to the same end. James Hill of the University of Colorado, as one example, is well known for attempting to discredit the relationship between obesity and higher insulin levels caused by sugar consumption. In his stated conflicts of interest, you will see that he has received consulting fees from Coca-Cola, Kraft Foods, and Mars Corporation (makers of Snickers, M&Ms, and Mars Bars).
Does this make him guilty of skewing data to serve corporate interests? No. But it should make you look closely at the studies themselves before accepting the headlines and making behavioral changes in your life.
The Goal of this Chapter vs. the Goal of this Book
Understanding how to act under conditions of incomplete information is the highest and most urgent human pursuit.
—Nassim Taleb, The Black Swan
Are the