Carbohydrate Restriction: How Low Should You Go?

A story about three studies and the evolution of our understanding of low carb diets.

When adopting a carbohydrate-restricted diet, many questions arise. Such as, how much of the deficit should be compensated? What ratio of fats to proteins is ideal?

These questions and more have been addressed in a few eloquent studies from which we can gather a lot of valuable insights.

Ketogenic Diet vs High Protein Diet

The first study (Johnston et al., 2006Trusted sourceNational Center for Biotechnology InformationNCBI's mission is to develop new information technologies to aid in the understanding of fundamental molecular and genetic processes that control health and disease.Open link) tested two high protein, low carb diets: one being ketogenic (10% carbs, 30% energy restriction); the other not ketogenic (40% carbs, 30% energy restriction).

Note that with enough of an energy restriction, 40% carbs could be ketogenic. On the other hand, 10% carbs may not be ketogenic in a large enough energy surplus. As you read these studies, make note of these details.

The study was fairly long (10 weeks) but only had about 10 participants in each group, mostly women with obesity.

The difference in fat loss was not statistically significant between the groups.

Also of note, at a strict 30% energy restriction, everyone is going to lose weight by definition.

Further, both groups cut carbs and junk food… this is what I call Hunger-Free Diet(s) because they didn’t mind only eating 70% of their usual fare. It turns out that when you give up industrial food, you don’t really need to replace the calories. Most industrial foods are high in refined carbs.

Ketogenic Diet vs Low-Carbohydrate Diet

The second study (Johnstone et al., 2008Trusted sourceNational Center for Biotechnology InformationNCBI's mission is to develop new information technologies to aid in the understanding of fundamental molecular and genetic processes that control health and disease.Open link) was shorter (two four-week phases), slightly larger (17 participants), but a crossover! In a crossover study, the participants are randomized to one of the diets for four weeks, then after a brief washout period, are switched to the other diet. In essence, in controls for any potential confounders.

Similar to the first study (Johnston et al., 2006Trusted sourceNational Center for Biotechnology InformationNCBI's mission is to develop new information technologies to aid in the understanding of fundamental molecular and genetic processes that control health and disease.Open link), both groups cut carbs to a degree, one down to 4%, the other 35%. Importantly, this study was ad lib. “Ad lib” means they could eat as much as they wanted.

Side note: it’s hard to make 4% carbs not ketogenic unless it was an energy surplus and a LOT of protein… which is pretty difficult, in part, because protein is so satiating.

Peculiarly this is what happened:

1) The ketogenic dieters didn’t like the food that much,

2) and causing them to eat less.

3) However, there was no big difference in weight loss (the ketogenic group lost slightly more weight than the low-carb group)

I attribute that oddity to having all of the food prepared by “kitchen research staff” LOL.

But the main point is that the degree of carb restriction didn’t matter much.

Just the fact of cutting out industrial foods worked in both groups. However, further reducing carbohydrate intake had no significant effect on weight loss.

Low-Carbohydrate Diet vs Low-Fat Diet

The last study is the DIETFITS Randomized Clinical Trial (Gardner et al., 2018Trusted sourceNational Center for Biotechnology InformationNCBI's mission is to develop new information technologies to aid in the understanding of fundamental molecular and genetic processes that control health and disease.Open link). This was the biggest and longest study of its kind. It was also ad lib and Hunger-Free Diet(s) (ie, they all spontaneously ate less).

Besides “clean-eating,” the low carb group was instructed to go as low as possible, which turned out to be ~30% carb (compared to 48% in the low fat group), whereas the low fat group cut fat down to 29% (compared to 45% in the low carb group). Both of the other studies were relatively high protein (30%, highly recommended), whereas this study was moderate in protein (about 22%).