When the low-carb hype doesn’t add up

The LA Times headline reads, "The case against carbohydrates gets stronger." The article is written by the lead author of a study recently published in the British Medical Journal (2018;363:k4583)

The study has received a flurry of attention from major news publications, so you've likely heard about it. Briefly: the authors put 120 people who had lost about 12% of their body weight on a maintenance diet for 20 weeks. The number of calories the participants consumed was carefully calculated (and adjusted, when necessary) for them to maintain their weight loss within about 2 kilograms (about 4.4 pounds). One-third of the participants were on a high-carbohydrate diet (60% of calories from carbohydrates), one-third on a moderate-carbohydrate diet (about 40% of calories from carbohydrates), and one-third on a low-carbohydrate diet (about 20% of calories from carbohydrates).

The authors used sophisticated methods to measure the total number of calories the participants burned, testing the participants on multiple occasions. Those who were on the low-carbohydrate diet appeared to burn more calories than those on a moderate- or high-carbohydrate diet, leading the authors to conclude that a low-carbohydrate diet would be more effective at helping people maintain weight loss.

The BMJ should be ashamed of itself for publishing this article. Here's what many reports on the study fail to mention:

First, the study actually begins with 234 participants in the weight loss phase. 70 of them (about 30%) failed to lose the required 10-12% of their body weight, leaving the authors with 164 people who began the test portion of the study - the maintenance phase. Forty-two (42) of those people failed to maintain their weight loss - that is, fully 25% of the participants failed to maintain their weight loss, despite the fact that the study provided the participants with all of their meals throughout both the weight loss and weight maintenance phases.

The authors base all of their analysis on only those who maintained their weight loss, completely ignoring the fact that 25% of their participants could not maintain their weight loss regardless of which diet they followed. This is what we in the research business call "cherry picking" - choosing to utilize only the data that supports your desired outcome.

Finally, if you look closely at the data, there's a plausible reason for why those on a low-carbohydrate diet burned more calories: before they started the weight loss phase they were more active than those in the moderate- or high-carbohydrate groups (in terms of physical activity), their total energy expenditure was higher than those in the moderate- or high-carbohydrate groups (just as they were after the maintenance phase, how interesting), and their lean body mass was higher and body fat mass lower.

What this means for you

This study is not just small, with only 146 people divided among the three arms of the study, it's simply throws out data on failure rates. It further fails to take into account meaningful pre-weight-loss differences between the groups.

What this study really proves is nothing we don't already know: losing weight by reducing calories doesn't work for everyone, and maintaining that weight loss is a challenge (but more physical activity helps). While losing weight is a useful goal, we no longer see weight loss alone as a meaningful goal for patients: it's far more important to improve the quality of the calories they put in their mouth, as that will have a far greater impact on their health risks than weight loss alone.

First posted: November 21, 2018