The trusted source for
healthcare information and
Altru Health System, Grand Forks, ND
Dr. Feldman reports no financial relationships relevant to this field of study.
SYNOPSIS: Researchers found that among adults seeking to maintain weight loss, consumption of a low-carbohydrate diet led to the highest total energy expenditure.
SOURCE: 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; Nov. 14;363:k4583.
Even today in our highly specialized, technologically advanced world of medicine, the centrality of diet to health is undisputed. However, what does remain in dispute are the specifics about dietary recommendations. Models and theories promoting a variety of dietary measures for disease treatment and control are numerous, and all too often studies have conflicting results.1 The lay press has published numerous articles proclaiming health benefits and successful weight control from diets as varied as the ketogenic diet, Whole30, and intermittent fasting. Yet, scientific data supporting such claims are lacking, and there are growing concerns about health problems, such as micronutrient deficiencies and cardiovascular safety problems, with long-term use of such diets.2,3
To bring some clarity to the field, Ebbeling et al conducted a randomized study with a narrow focus on energy expenditure varying with diet during maintenance of weight loss. The group did not look at the weight loss phase, did not use weight as an outcome measure, and did not look at any health outcomes beyond the 20-week study period. The primary outcome was a measure of energy expenditure. Secondary outcomes were differences in physical activity and metabolic hormones. All patients effectively completing the study maintained weight via a process of diet adjustment. Ebbeling et al stated that their work contributes to the understanding of metabolic variations induced by diet, and could lead to recommendations essential in the treatment of obesity.
An understanding of the carbohydrate-insulin model of obesity is useful to put this study in perspective. Basically, this model looks at insulin levels following ingestion of high carbohydrate load. The model suggests that higher ratios of insulin to glucagon lead to greater storage of fuel in fat tissue and that this state increases hunger and cravings for food and lowers energy expenditure, especially in persons with constitutionally high insulin secretion. However, controlled studies have not supported this model.4
Energy expenditure refers to the overall effect of energy required for basal metabolism (metabolic functions while at rest) plus the energy needed to absorb and digest food (thermic effect) plus energy required for activity. In simple terms, weight gain occurs when energy from food is greater than the total energy expenditure. For weight maintenance, balance among these components is necessary.5
This study began with a 10-week “run-in phase” during which time participants were monitored closely, provided with meals, and given an activity level to promote weight loss of 10-14%. All participants were between the ages of 18 and 64 years and had a starting body mass index > 25 kg/m2. Participants were recruited from several college campuses and were given stipends for participation. Those who achieved the targeted weight loss were eligible for participation in the randomized potion of the study.
The researchers randomized 162 participants to either a high-, medium-, or low-carbohydrate diet for 20 weeks. For all three diets, the protein level was fixed at 20% of total energy. Carbohydrate and fat content varied inversely; the high-carbohydrate diet contained 60% of energy from carbohydrates and 20% from fat; the moderate-carbohydrate diet contained 40% of energy from carbohydrates and 40% from fat; and the low-carbohydrate diet contained 20% of energy from carbohydrates and 60% from fat.
Energy expenditure (the sum of basal metabolic rate, activity level, and thermic effect of food) was calculated via analysis of urine samples after administration of doubly labeled water. This technique for measuring energy expenditure has been used to measure energy expenditure in humans since the 1980s. The process involves substituting an isotope for oxygen and hydrogen in water molecules, and then examining urine samples to determine average metabolic rates.6
Ebbeling et al reported two sets of data: intention-to-treat group (all participants post “run-in phase”; n = 162) and per-protocol group (data only from the participants who maintained weight during the 20-week test phase and completed all study requirements; n = 120). Table 1 shows energy expenditure measured by doubly labeled water for each specified carbohydrate diet level (high, medium, or low) in the intention-to-treat group and the per protocol subgroup.
Secondary outcome measures included time spent in physical activity. There was no significant difference among the groups in this area, nor was there a difference in time spent in sedentary activities. Ghrelin, known as the “hunger hormone,” was significantly lower among those in the low carbohydrate diet.
To determine if high insulin secretors were more vulnerable to changes in energy expenditure linked to variations in carbohydrate load, each participant was assessed at baseline for fasting glucose and serum level of insulin secretion after oral glucose administration. Three groups were determined (high, medium, and low insulin secretors) based on response. Table 2 shows each of these three insulin secretor groups and compares energy expenditure in each across high- to low-carbohydrate diets (includes the intention-to-treat and per-protocol groups).
The popular press seemed to embrace this study. The “Today” Show proclaimed: “For weight loss maintenance, a low-carb diet may be best.”7 CNN and other media outlets followed with similar themes.8,9 However, behind the headlines, significant questions from the scientific community remain regarding methodology and applicability to the general population. To be fair, Ebbeling et al explicitly noted in their conclusion that
additional research is necessary to confirm conclusions and generate recommendations for the public. Perhaps the most consistent criticism in the literature regarding the study is the use of doubly labeled water as the measure of energy expenditure. It is thought that the changes noted in measurements of energy expenditure may reflect an artifact of methodology due to differences in fat oxidation under conditions of high-fat and low-carbohydrate diets.10 Although Ebbeling et al did not agree with this particular point, they noted that relying on this method to measure energy expenditure has definite limitations and requires assumptions that may not be completely accurate.
Notably, relying on any measures of energy expenditure rather than actual weight change makes it difficult to translate the results into usable recommendations for patients or the public. In the study, efforts were made to maintain weight loss throughout the 20-week test phase. These efforts are described as “adjustment of energy intake to achieve weight stability.” However, if the low-carbohydrate diet had the highest rate of energy expenditure, it may be expected that this group would either lose weight during the 20 weeks or ingest more calories to prevent weight loss. As there was not a significant difference in weight change among any of the groups at the conclusion of the study, it would be useful to know if, in fact, the low-carbohydrate group consumed more calories. This may be an avenue for future investigation and would be helpful to translate the findings to usable recommendations.
Ebbeling et al noted there were significant efforts to ensure dietary compliance throughout the study, including supplying all participants with prepared meals and WiFi scales and performing spot checks for diet and weight compliance. However, even with these efforts, compliance and weight accuracy cannot be guaranteed; future investigators may be able to implement newer technologies to determine accuracy and compliance.
On a related note, a major barrier to applying findings from this study to the general population is the need to obtain dietary compliance. Following a strict low-carbohydrate diet could be challenging for many in the obese population. Additionally, since the low-carbohydrate diet included a high percentage of calories from fat (60%), from a health prospective it is imperative to make certain that following such a diet does not introduce unintended health risks.
Finally, as with many preliminary studies, several unknown or unmeasured factors need identification over time. As previously alluded, the diets differed in both fat and carbohydrate levels. Although the proportion of saturated to unsaturated fat was consistent in all groups, the actual amount of fat varied. Because of the need to change levels of macronutrients, the foods in the diets were different. Both factors — varying amounts of fat and food differences — may have introduced variables that affected the study results.
The take-home message from this study is that there is suggestive evidence that diet can influence metabolism, and that a low-carbohydrate diet may be best for increasing metabolic energy, especially in high insulin secretors. There is no evidence at this point that these findings translate seamlessly to dietary recommendations for weight loss, but suggestive evidence for patients wanting weight-loss maintenance.
Finally, it is reasonable to remember that an effective diet depends on patient acceptance, understanding, and willingness to implement. Dietary compliance is a multifactorial process determined on many levels with not only medical influences, but also societal, cultural, and economic influences. Developing an understanding of these and other factors with each patient will allow clinicians to create an individualized diet as part of a holistic treatment plan.
Financial Disclosure: Integrative Medicine Alert’s Executive Editor David Kiefer, MD; Peer Reviewer Suhani Bora, MD; Relias Media Executive Editor Leslie Coplin; Editor Jonathan Springston; and Editorial Group Manager Terrey L. Hatcher report no financial relationships relevant to this field of study.