Channeling Nina Teicholz: Latest LC/LF Study Should Be Ignored

I'm sure by now you've heard ... low carb "bested" low fat once again in a gold standard randomized controlled clinical trial.  As usual, some weighed in even before the full article was published, but I guess this is to be expected these days.

Before discussing this study, perhaps next week, I wanted to point out that by the standards of Big Fat Surprise author Nina Teicholz, this study should be summarily dismissed.  Why?  Well:

Not Representative of the General Population

It was conducted in 148 obese adults of which 88% were female and 51% were black. Thus the results are not applicable to the general population. Teicholz relegated a well designed and implemented 2 year RCT on the Mediterranean diet to a footnote in her book for this reason. I discussed this in detail here, which was prompted by this footnote in BFS:

XXXI . There was one other long-term (two-year) trial comparing a low-fat to a Mediterranean diet, with results in 2004. It showed that the Mediterranean diet performed better. But the study involved men and women with metabolic syndrome, not a normal population, and could not be generalized (Esposito et al. 2004). {Kindle Location 3493/12033}

You can click on the table to enlarge if you don't want to view the blog post.

Clearly a trial conducted on a group with metabolic syndrome would not be generalizable to the American population - [/sarcasm]. I don't understand why large trials such as this are conducted if they aren't going to even attempt to make them representative of the general population, or focus on a specific sub-population. You want to look at women? Then do your study in all women -- why throw 12% men in the mix when there are clear gender differences in the way metabolic parameters such as abdominal fat and HDL are concerned (to name just two). In the study they mention wanting to see how the diet impacted blacks who may have been underrepresented. Again, given as body composition was a factor, and there are some known racial differences, then do a study only in blacks. Was the distribution of race equal amongst the genders? We are not provided with that information and there is no indication that any post hoc analysis was done with subsets of subjects to control for either factor.

Race factored prominently into the Discussion section of this paper.  The subjects "included a substantial sample of black persons (a group underrepresented in previous trials)".  Yet we can all agree that 51% would be a gross overrepresentation.  I wonder why they didn't present a separate analysis for just the blacks, even if just to say that they did so and found no differences. 

It was Not "Controlled"!!

By that I mean, using Teicholz's definition regarding compliance. Teicholz has brutally lambasted the LA Veterans study, both in the book and moreso even in some interviews. I discussed this in depth here. Teicholz declares:

In the best (most “well-controlled”) trials, researchers prepare or provide food to study participants to control exactly what they eat. Sometimes subjects are invited to dine in a special cafeteria, or sometimes the researchers will go so far as to deliver meals to their subjects ’ homes— although these kinds of measures can be quite expensive. In less well-controlled trials, subjects are simply counseled about what to eat and perhaps given a diet book to take home. {1297}

As it turns out in this trial, unlike the LA Vet trial where special meals were prepared for live-in residents, 3 meals per day ... and unlike even the Shai trial where special labeling of meals in the workplace cafeteria was used for 5 self-selected meals per week ... there was no food prep. The participants were provided with one diet appropriate meal replacement shake/bar per for the duration of the trial (brand/source/funding unspecified).  Other than that, as Teicholz would say, we have no idea what they were eating!   Did they eat the fakey foods?

 The participants filled out two 24-hr recalls at each time point (one week/one weekend day, baseline/3/6/12mos).  There is NO -- as in zero, zilch, nada -- description of the nature of the diets.  

They Didn't Control for Smoking ...

... or even account for it.