Sunday, January 4, 2009

Consumer Reports “Secrets of Thin People”

The February 2009 issue of Consumer Reports features an article called “Dieting on a budget, plus the secrets of thin people, based on our survey of 21,000 readers.” The feature includes some equipment reviews, plus articles on “healthful eating on a budget” and their analysis of their reader survey results.

I often find that I disagree with articles and reviews in Consumer Reports, more so whenever I actually have strong personal knowledge of a product or subject area. Their articles are not, of course, peer-reviewed scientific reports. Nevertheless, and despite the fact that they carefully avoid accepting commercial advertising and sponsorship, they typically conform to current industry norms, assuming, I guess, that their readers are responding to the advertising and product offerings they are exposed to and are merely wanting to choose which of the available highly-promoted products to buy. The same can be said of their financial and medical advice in that their advisors (and authors??) seem to be drawn from the mainstream establishment.

There was a time, a few decades ago, when product reviews often used a home-made recipe or similar non-commercial product as a reference for comparison. Quite often, the commercial products did not compete favorably with the non-commercial version. Such comparisons are very rare nowadays. I suppose it, at least partly, reflects the fact that fewer and fewer people are inclined to make their own anything, even when the results are superior. (Though I will admit that there are at least some real examples where modern manufacturing methods make commercial products that are superior…)

Anyway, I digress. In typical fashion, the results of the survey are presented in a way that contains almost no real data and certainly no statistical analysis. Here's what I think are the real “data”:

Survey date: 2007; 21,632 respondents; 66% “overweight” including 22% “obese”
  • 16% “never overweight”
  • 15% “successful losers” (defined as weighing at least 10% less than at their heaviest and keeping the weight off for at least 3 years)
  • 42% “failed dieters”
  • 27% other
“more than half” of successful losers did so without the aid of a commercial diet program (defined as a particular medical treatment, book, or pills)

The data get pretty thin after that! The report claims that “through statistical analyses [not presented], we were able to identify six key behaviors that correlated most strongly with having a healthy body mass index.” The standard threshold of 25 is used for defining “overweight,” even though such arbitrary BMI thresholds are well-known to be misleading due to variability in skeletal frame size and muscle development.

The basic conclusion of the study was that successful losers generally embraced the allegedly good behaviors naturally practiced by the always-thin, typically slightly more so, and “significantly” more so than did the failed dieters. Therefore, the authors conclude, all you need to do to be a successful loser is to “quite literally, live like a thin person.”

The six behaviors (“secrets of the slim”) recommended turn out to be current mainstream party-line recommendations:

 Succ. losersAlways thinfailed dieters
watch portions62%57%42%
limit fat534735
eat fruits and vegetables494938
eat whole grains“consistently opted for”?
eat at homeweight correlated with number of meals out
“exercise, exercise, exercise”*323123

*while the authors suggested that vigorous exercise that increased breathing and heart rate for 30 minutes or longer was strongly linked to lower BMI, the survey numbers are for strength training at least once per week

The authors note that going low carb is conspicuously absent from the list and that limiting carbohydrates correlated with higher BMI in the survey, though they did note that such a correlation could be an artifact of the fact that those with higher starting BMI are more likely to try a low carb diet. Despite this admission, they still claim that “the findings do suggest that cutting carbs alone, without other healthful behaviors such as exercise and portion control, might not lead to great results.”

They also reported that three other strategies did not show significant effects: eating small meals, never eating between meals. and including lean protein with most meals.

Surveys like this are notoriously unreliable for establishing any true cause-and-effect relationships. It is easy to bias the questions to support the conclusion you want to reach. The conclusion about the ineffectiveness of low carb diets is a good example. If only a small percentage of the survey population had even tried it, then, of course, such a diet would show low significance overall, and therefore no conclusion whatsoever can really be drawn as to the effectiveness of a low carb diet.

Another aspect of the report is the definition of “successful” dieters as those who’ve achieved a 10% weight loss (and kept it off for at least 3 years). For the 22% of respondents characterized as “obese,” 10% is barely a good starting point. At the end of the article, the authors, in a section labeled “Realistic goals are one key to weight loss,” state, “A 10 percent loss might not sound like much, but it can significantly improve overall health and reduce risk of disease.” True perhaps, but pretty discouraging from a public health point of view, if that’s the best we can hope for! And we clearly now know better. There are countless examples of people who have successfully lost much more weight and kept it off (for example, by carbohydrate restriction for those who have demonstrated sensitivity to carbohydrate consumption), and they’ve done it without resorting to extreme methods such as bariatric surgery or starvation diets.

At best, one can use survey data like this to construct hypotheses that might be worth testing. To base public health recommendations on the results is ridiculous!

I look at the data and draw a rather different set of conclusions:
  1. We have a serious weight problem if 5 out of 6 people cannot be identified as always thin!
  2. Current recommendations don’t work!
  3. Even those mainstream recommendations that maybe seem to help don’t help very much.

And I would tend to characterize most of the six “secrets of the slim” as not secrets at all, but just things that are sometimes associated with the habits of people who successfully control weight, and not things that can fix it. Clearly, you can do all six “right” and not lose weight. Yes, overeating (oversized portions, eating out a lot) can make you fat, but it’s much more important what you eat and how you manage what you eat: control of portion size is realistically about controlling hunger and satiety and not about counting calories.

The reported survey benefits of lower fat, higher fruits, vegetables, and whole grains probably just correlate with respondents who are generally attempting to pursue healthful habits, a population who are more likely to exercise self control at earlier signs of a weight problem anyway. Also, the statistics presented do not make a very convincing case that these particular food choices are significantly beneficial; only about half of each survey group practiced any particular allegedly beneficial choice anyway. Hard evidence of benefits of such dietary recommendations are actually thin to non-existent. Exercise is good for overall health, but it's a poor strategy for weight control. Being completely sedentary is correlated with obesity, but it is not a cause. If anything, it’s the other way around: obesity causes people to lead a sedentary lifestyle.

And indeed, one could make a similar argument about causation with respect to the whole survey: always-thin people tend to eat less and exercise more because their body condition makes them so inclined. They are sated with less food and find exercise easier and more enjoyable. Trying to emulate the effect (eating less and exercising more) to achieve the cause (weighing less) doesn't work very well! The authors’ conclusion that all you need to do to be a successful loser is to “quite literally, live like a thin person” gets cause and effect backwards.

1 comment:

Anonymous said...

I recently came across your blog and have been reading along. I thought I would leave my first comment. I don't know what to say except that I have enjoyed reading. Nice blog. I will keep visiting this blog very often.