Do low carb diets increase energy expenditure? [Study review]

The fats vs. carbs debates seem to be never ending and a new study just added fuel to the fire. Ebbeling et al. (2018) conducted a massive study on the metabolic effects of high vs. low carb diets. 1685 people were screened and 234 of those were offered a participation ‘salary’ of $3280 on top of getting $3220 worth of prepared meals to participate in a 16-week weight loss diet followed by a 20-week study period. Got money?

Their main conclusion was: “In this controlled feeding trial over 20 weeks, we found that total energy expenditure was significantly greater in participants assigned to a low carbohydrate diet compared with high carbohydrate diet of similar protein content. In addition, pre-weight loss insulin secretion might modify individual response to this diet effect.”

The study compared the metabolic effects of 20%, 40% and 60% carb diets at weight maintenance in obese individuals. Actual weight maintenance, not estimated. Protein intake was 20% in all groups and sugar intake was 15% in all groups. It was a free-living study, but meals were provided and several measures were taken to ensure compliance. More on this in a bit. The main result was that in spite of consuming the same protein and energy intake, energy expenditure differed by several hundred calories per group. The lower the carbohydrate intake, the higher the energy expenditure. The effect was stronger in people with greater symptoms of metabolic syndrome, specifically greater pre-weight loss insulin secretion after an oral glucose test. When comparing the most diabetic individuals, the 20% carb diet resulted in almost 500 calories more energy expenditure than the 60% carb diet.

I’ve seen 2 reactions to these results.

1. Case closed, low carbohydrate diets all the way! This roughly seems to be the direction the authors veer in.
2. The authors are all keto zealots, so these results are BS. A calorie is a calorie.

A calorie is indeed a calorie. That’s tautological. However, that does not mean a given energy intake always has the same effect on your body composition. While most diets with the same energy and protein intake result in similar fat loss, several studies have found that certain populations, especially carb intolerant individuals presumably like those in this study, lose less fat on high glycemic load diets.

• The same gross energy intake may not have the same digestible or metabolizable energy intake.
  Example: certain types of fiber cannot be digested or metabolized by the human body.

The food which can be digested and metabolized can then have different metabolic effects.

• It can affect energy expenditure by thermogenesis, basically the energy it takes to digest, absorb and metabolize the food.
  Example: medium-chain triglycerides are well known to have a greater thermic effect than other saturated fatty acids due to its different metabolism.
• It can affect spontaneous physical activity by modulating the autonomic nervous system and by psychological factors. Feeling more energetic may cause you to move more.
• It can affect nutrient partitioning, resulting in greater energy expenditure due to the building or retention of lean body mass.
  Example: protein intake stimulates muscle protein synthesis, which increases energy expenditure.
• Somewhat more hypothetically, different foods can have different effects on brown fat thermogenesis or metabolic efficiency.

In short, there are numerous ways your diet can affect your energy expenditure, so the current study’s results are not impossible. The study actually has good information about which factors likely caused the difference in energy expenditure.

The first thing I looked at, like in any free-living study, was non-compliance. If the lower carb groups ate more, they had a higher diet-induced thermogenesis and this could explain the greater energy expenditure. While the authors used several strategies to improve compliance, including providing meals, energy intake did indeed differ with the same trend energy expenditure did and the difference reached statistical significance in the people with the highest difference in energy expenditure: “energy intake changed in participants assigned to the high, moderate, and low carbohydrate test diets, respectively: 139 kcal/d, 175 kcal/d, and 269 kcal/d, with an overall P=0.36. These differences strengthened among participants in the highest third of insulin secretion: 37 kcal/d, ?24 kcal/d and 340 kcal/d, with an overall P=0.05.”

However, this could not explain the results. The thermic effect of food (TEF) in a population like this is generally 10%, less even often when protein intake is kept constant. To explain several hundred calories worth of energy expenditure, the difference in energy intake would thus have to be several thousand calories.

It’s implausible they ate that much more, as the low carb groups experienced greater reductions in the hunger hormone ghrelin and lesser decline in the appetite stimulating hormone leptin.

Plus, the participants were all weight stable, so they couldn’t have had huge differences in energy intake. All in all, TEF from overeating could at best explain a fraction of the results.

What could then explain the difference in energy expenditure? The most obvious marker to look at is basal metabolic rate (BMR). Many low carb proponents claim a low carb diet stimulates BMR. However, resting energy expenditure actually did not differ across groups in this study. If not overeating and not BMR, that basically leaves activity. Indeed, moderate-to-vigorous activity decreased significantly more in the moderate and high carb groups. However, this again couldn’t explain several hundred calories worth of energy expenditure, as the difference was only a few minutes per day. The decrease was also largest in the medium-carb group, not the high-carb group, and there was no significant difference in total sedentary time or total physical activity step count. Skeletal muscle work efficiency also did not differ per group.

But what if we count everything, including all the non-significant differences, together? We have in the total sample 278 calories greater energy expenditure in the low vs. high carb group. Minus an expected 0.1 x 340 = 34 kcal from overeating TEF (based on maximum energy intake divergence), ~33 kcal from the non-significantly different physical activity level (which is generous based on the trivial difference in step counts) and 100 kg * 0.1 kcal/kg/min * 5 minutes = 50 kcal from the increased moderate activity level, that leaves 161 calories to be accounted for by factors like TEF and nutrient partitioning. That’s plausible but also within the realm of measurement error and non-compliance.

In conclusion, I’d say the general take-home message of this study is that obese individuals, especially ones on the verge of type II diabetes, don’t fare well on highly glycemic diets (read: starches and other processed carbs). Not because of major metabolic differences, but because they’ll likely be hungrier (more ghrelin, less leptin) and they anecdotally don’t feel as energetic, which may unintentionally suppress their physical activity level. The inherent metabolic differences as a result of higher vs. lower carb intakes shouldn’t amount to large differences in energy expenditure in lean, muscular strength trainees with good carb tolerance, so if you strongly prefer high carb diets, this study is probably not a reason to change your diet.

New study reference

Effects of a low carbohydrate diet on energy expenditure during weight loss maintenance: randomized trial. Ebbeling et al. BMJ 2018; 363 doi: https://doi.org/10.1136/bmj.k4583 BMJ 2018;363:k4583