Ebbeling CB, Feldman HA, Klein GL, et al. Effects of a low carbohydrate diet on energy expenditure during weight loss maintenance: randomized trial. BMJ. 2018;363:k4583.
Randomized controlled trial that featured a run-in weight loss phase.
To measure the effects of diets with varying ratios of carbohydrate to fat on total energy expenditure.
This trial recruited a total of 164 adults aged 18 to 65 with a BMI of 25 or more.
After a run-in phase of 12% (10%-14%) weight loss, participants were randomly assigned to receive a low-carbohydrate (20% of calories), moderate-carbohydrate (40% of calories), or high-carbohydrate (60% of calories) diet for 20 weeks.
Diets were adjusted calorically to maintain weight loss within 2 kg throughout the 20-week intervention. Protein intake was a controlled amount at 20% of caloric intake. Fat made up the residual 60%, 40%, or 20% of total calories for each participant. Meals were distributed to the participants, who were required to weigh themselves daily at home and transmit the data to the investigators electronically.
Study Parameters Assessed
Assessments included (but were not limited to) body weight; energy intake; physical activity; total energy expenditure, calculated using estimated CO2 production; ghrelin, leptin, and postprandial insulin secretion before weight loss.
Primary Outcome Measures
The primary outcome was total energy expenditure and secondary outcomes included ghrelin and leptin levels.
There was a linear trend of increasing total energy expenditure (52 kcal/d) for every 10% decrease in proportion of carbohydrate intake (P=0.002). Compared to the high-carbohydrate cohort, those in the moderate and low cohorts had higher total energy expenditure of 91 kcal/d and 209 kcal/d on average.
This effect was greater in those who started with high insulin levels at baseline. In those in the highest third of pre–weight loss insulin secretion, there was a 308 kcal/d difference between the low- and high-carbohydrate diets in the intention-to-treat analysis and a 478 kcal/d in the per protocol analysis (P<0.004).
Both ghrelin and leptin were significantly lower in the low-carb group with ghrelin showing a steeper decline than leptin.
This study is the largest and longest randomized trial published to evaluate the weight loss effects of a low-carbohydrate diet. It suggests that not all calories are created equal and implies that the current obesity epidemic may not be attributable to fat intake or excess calories but, more specifically, to excess carbohydrates.
Key strengths of this study are the study duration and ability of the researchers to control food intake. In a previous study comparing low-fat with low-carb diets, Gardner et al found no significant difference in weight loss; however, that intervention used merely health education with no control over actual food intake.1 A recent meta-analysis by Hall and Guo showed no significant difference in energy expenditure between low-carb and high-fat diets.2 This was likely due to the fact that the studies featured in that analysis were short-term, mostly less than 2 weeks in duration. The authors of the current study under review believe it may take at least 2 or 3 weeks to adapt to a low-carbohydrate, high fat-diet.
While the primary outcome of weight loss due to increased caloric burn is interesting, the secondary outcome (ghrelin and leptin) provides additional information on why low-carb eating may lead to weight loss even when it is isocaloric.
It’s also interesting that the low-carbohydrate cohort had a moderate increase in basal metabolic rate and increased physical activity, which contributes to the long-lasting effects of low-carb eating when it comes to weight loss and maintaining normal body weight.
While the primary outcome of weight loss due to increased caloric burn is interesting, the secondary outcome (ghrelin and leptin) provides additional information on why low-carb eating may lead to weight loss even when it is isocaloric. Ghrelin is a hormone that helps regulate appetite and lipid metabolism. When it is elevated, it increases hunger and food intake.3 Ghrelin also plays a critical role in insulin metabolism, with increased levels contributing not only to obesity but also metabolic syndrome and type 2 diabetes.4 Leptin is also a hormone that regulates appetite and energy balance and studies indicate that obese individuals have increased leptin concentrations.5 Both ghrelin and leptin were significantly lower in the low-carbohydrate cohort of the current study.
The ghrelin/leptin connection between low-carb eating and weight loss explains an important aspect of weight loss, which is appetite control. Enhanced satiety results in the consumption of fewer calories. Hu et al demonstrated that a low-carb diet increased peptide YY, a satiety signal, when compared to a low-fat diet, which demonstrated that there was better appetite control in the low-carb group.6
Defining a low-carb diet can be tricky. A 2013 paper published in the journal Current Diabetes Reports defines the carbohydrate spectrum in this manner (based on a 2,000 calorie diet):7
- Very low = 21 to 70 g/day (4.2%-14% of caloric intake)
- Moderately low = 150 to 200 g/day (30%-40%)
- Moderate = 200 to 325 g/day (40%-65%)
- High = >325 g/day (>65%)
A review published in the Journal of the American Osteopathic Association makes these distinctions:8
- Low-carb/high-fat = 20 to 100 g/day
- Ketogenic = <50 g/day
- Atkins = induction <20 g/day then 80 to 100 g/day
- Zone = 40% of total calories
In the present study, carbohydrates comprised the following percentage of total calories:
- High = 60%
- Moderate = 40%
- Low = 20%
Because of the inconsistent definition associated with “low-carb,” studies on the outcomes of low-carb diets can be challenging to compare. Research involving the ketogenic diet, which has the lowest levels of carbohydrate intake, has been increasing, especially in the areas of cardiology9 and oncology.10 Poor patient compliance and overconsumption of unhealthy fats are problems associated with the ketogenic diet and other ultra-low–carb diets. Poor food quality choices may skew results in those studies, and may not be reflected in the data if not specifically measured as a confounder. Tracking of micronutrient intake, proinflammatory food consumption, processed foods, and subtypes of macronutrients (ie, fermentable vs unfermentable carbs) would help parse the data and lead to more rigorous conclusions. For now, it’s important for practitioners to direct their patients away from consuming pro-inflammatory foods and towards healthy food choices regardless of the underlying diet type.
While a low-carb diet has been shown to benefit heart health,11 a recent observational study called that into question. The study was published this year in the Journal of the American College of Cardiology and followed 13,852 participants (average age of 54, 45% male) over a median of 22.4 years. They found there was a correlation between a low-carb diet (<39% of calories) and an increase in atrial fibrillation.12 However, there are several issues that make the results of this study questionable. First of all, 39% carbs in the diet is not considered low-carb (as described previously). The study also used a food frequency questionnaire, which can be unreliable. Lastly, there appeared to be a U-shaped curve, where the group consuming the most carbs experienced a 16% increased risk of atrial fibrillation as well. Furthermore, confounding factors such as hypertension, diabetes, and other risks for atrial fibrillation were not considered.
While that study received some significant media attention, more research in the form of randomized clinical trials is needed before advising patients. In fact, other research has indicated that a low-carb diet can help reduce the risk of conditions that increase risk of atrial fibrillation, such as hypertension and diabetes.13-15
We agree with Ludwig and Ebbeling, who wrote, “Despite intensive research, the causes of the obesity epidemic remain incompletely understood and conventional calorie-restricted diets continue to lack long-term efficacy…Pending definitive studies, the principles of a low-glycemic load diet offer a practical alternative to the conventional focus on dietary fat and calorie restriction.”16 This latest study provides further evidence that the advice to simply eat less and burn more is outdated.
Integrative practitioners who employ dietary advice as a key intervention will likely want to consider helping patients eat a low-carb diet to achieve lasting weight loss and, in the process, reduce risk of chronic illness. According to the study reviewed here, patients with the highest levels of pre–weight loss insulin may benefit the most from this (low-carb diet) as a lifestyle choice.