I like Chris, he's one of the very few serious nutrition scientists.
In the HLF vs HLC diets, respectively, the mean 12-month macronutrient distributions were 48% vs 30% for carbohydrates, 29% vs 45% for fat, and 21% vs 23% for protein. Weight change at 12 months was −5.3 kg for the HLF diet vs −6.0 kg for the HLC diet (mean between-group difference, 0.7 kg [95% CI, −0.2 to 1.6 kg]).
Quoted from the abstract. I see a tendency here, a significant tendency that replicates pretty much all other diet experiments. The more fat, the better the results. The less carbs, the better the results. And, protein is inconsequential because it's very hard to go very high or very low, which would be needed to produce any actual significant difference.
(In further reading the full paper, there's a problem with the idea of more fat. The HLC group actually ate less fat too. Without knowing any more about the type of fat they ate (which is arguably pertinent for various reasons), I'm left with the idea of less carbs. The HLC group cut their carb intake by half. This makes the "non-significant" results even more puzzling to me.)
See, Chris concludes there's no significant difference in the results, but outside the context of the equally weak intervention difference. Once we consider the weak intervention difference, results become more significant. Of note, the tendency I mentioned was not reversed. If it had, I'd wonder how that's even possible.
One thing just struck me. If we accept Chris' conclusion of "no significant difference". It does not mirror Chris' own A-TO-Z experiment in terms of intervention-results correlation. Tendency is still there, but somehow made either weaker for the HLC group, or stronger for the HLF group, or both. In the A-TO-Z experiment, there is a clear difference in power of intervention when considering amplitude of results. Atkins was best - more powerful - simple as that. But here, if we equate HLC to Atkins, it's no longer best, it's merely equivalent.
Both groups were not advised to eat less total food, except specifically for eating less fat or eating less carbs. There was no intentional total calorie restriction. This could be a pertinent finding so that regardless of which diet we choose for weight loss, eating to satiety somehow has an effect of its own. It makes a certain sense if we consider the Minnesota semi-starvation experiment, where subjects went crazy with hunger. If this effect is true, that alone contradicts official guidelines regarding weight loss.
All in all, this latest from Chris is weaker and less reliable than the A-TO-Z experiment. In a way it makes sense. I had the impression that the idea of genotype/insulin-sensitivity was a result of the A-TO-Z results in that regard. That experiment was more powerful, more reliable, less ambiguous. This means to me that the DIETFITS experiment can't confirm nor refute this idea simply because it's too weak to do so. I feel like Chris tried to reinvent the wheel, when it looks like he just figured out how to lose a wheel. Atkins got it right the first time, stick to what works, Chris.
Here's a thought. The idea of genotype/insulin-sensitivy is that it's all about our genes, and the effect of diet on those genes. Well, if the idea includes the concept of an effect from diet, this is an environmental factor, so it invariably includes the concept of other environmental factors as well. As Taubes said, carbs is the primary
cause, not the only
cause. If carbs have the ability to affect genes, then other environmental factors have it too, to make carbs (or other macros or whatever) appear more or less powerful. This means that for an experiment to expose this other factor, carbs must be removed, not just cut by half. In my opinion, it explains why DIETFITS can't show what's going on with genotype/insulin-sensitivity.