Low-carbohydrate diets lead to greater weight loss and better glucose homeostasis than exercise: a randomized clinical trial

Abstract

Lifestyle interventions, including dietary adjustments and exercise, are important for obesity management. This study enrolled adults with overweight or obesity to explore whether either low-carbohydrate diet (LCD) or exercise is more effective in metabolism improvement. Forty-five eligible subjects were randomly divided into an LCD group (n=22) and an exercise group (EX, n=23). The subjects either adopted LCD (carbohydrate intake<50 g/day) or performed moderate-to-vigorous exercise (≥30 min/day) for 3 weeks. After the interventions, LCD led to a larger weight loss than EX (−3.56±0.37 kg vs. −1.24±0.39 kg, P<0.001), as well as a larger reduction in fat mass (−2.10±0.18 kg vs. −1.25±0.24 kg, P=0.007) and waist circumference (−5.25±0.52 cm vs. −3.45±0.38 cm, P=0.008). Both interventions reduced visceral and subcutaneous fat and improved liver steatosis and insulin resistance. Triglycerides decreased in both two groups, whereas low-density lipoprotein cholesterol increased in the LCD group but decreased in the EX group. Various glycemic parameters, including serum glycated albumin, mean sensor glucose, coefficient of variability (CV), and largest amplitude of glycemic excursions, substantially declined in the LCD group. Only CV slightly decreased after exercise. This pilot study suggested that the effects of LCD and exercise are similar in alleviating liver steatosis and insulin resistance. Compared with exercise, LCD might be more efficient for weight loss and glucose homeostasis in people with obesity.

Keywords

low-carbohydrate diet / obesity / nonalcoholic fatty liver disease / continuous glucose monitoring / mean sensor glucose

1 Introduction

The rising pandemic of obesity has become a global public health concern. According to the Global Burden of Disease study in 2015, the overall prevalence of obesity among children and adults was 5.0% and 12.0%, respectively; obesity increases the incidence and mortality of many chronic noncommunicable diseases and imposes enormous social and economic burdens on individuals, communities, and countries [1].

Lifestyle interventions, including dietary adjustments and exercise, are considered as important methods for weight loss and the management of various diseases. Low-carbohydrate diets (LCDs) have recently regained popularity in the public and scientific communities. Energy surplus, excessive consumption of sugar, and lack of exercise could lead to obesity and type 2 diabetes mellitus (T2DM). Clinical studies indicated that LCD has a positive influence on body mass index (BMI), waist circumference, blood pressure, serum lipid levels, and glucose metabolism [2,3]. Physical exercise is another effective way to control body weight and improve metabolism. Moderate and vigorous exercise can reportedly reduce intrahepatic triglycerides, body weight, waist circumference, and blood pressure [4,5].

Healthcare professionals and researchers generally recommend that diet modification should be combined with regular physical training during weight control. However, people with obesity usually tend to choose only one of them because of their unwillingness to change their eating habits or lack of time for exercise. In addition, the exact difference between LCDs and exercise in metabolism is unclear. Identifying the respective influence of various confounding factors on weight loss would be difficult if the two interventions are simultaneously conducted. Therefore, a head-to-head comparative trial involving LCDs and exercise was designed to guide clinical practice.

Continuous glucose monitoring (CGM) devices are subcutaneous implantable electrochemical glucose enzyme-based sensors that can continuously monitor glucose levels for several days. CGM devices are now widely used in monitoring people with diabetes mellitus [6]. CGM devices can reflect glycemic fluctuations and are a useful tool for visualizing the effects of diets, exercise, and medications on glucose. However, only a few studies used CGM devices to investigate the effects of LCDs and exercise on people without diabetes during the entire study period.

The current study aimed to compare the 3-week effects of LCDs and exercise on weight loss and metabolic parameters in nondiabetic adults with overweight or obesity. Their fat content was evaluated via magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS). CGM sensors were applied during entire trial period to continuously record their glucose levels. Thus, this study provides new insights into scientific weight loss intervention.

2 Material and methods

2.1 Study population

The study recruited subjects with overweight or obesity from the outpatient clinic at the Department of Endocrinology and Metabolism of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital from July 2018 to January 2019. The inclusion criteria were as follows: no history of diabetes mellitus, aged 18–40 years, BMI≥24 kg/m², waist circumference of males≥90 cm or females≥85 cm [7], and a sedentary lifestyle (>3 h/day sitting and<150 min/week of moderate-intensity exercise for>3 months). The exclusion criteria were as follows: fasting plasma glucose (FPG)≥7.0 mmol/L or 2 h plasma glucose (2hPG) after a 75 g oral glucose tolerance test (OGTT)≥11.1 mmol/L or glycated hemoglobin A_1c (HbA_1c)≥6.5%; pregnancy or lactation; severe liver, kidney, or thyroid dysfunction; dyspepsia or malabsorption; history of malignancy or mental disorders or coronary vessel diseases; use of medication that might influence metabolism; history of alcoholism or drug abuse; history of metabolic surgery; participation in a weight loss intervention within 1 month before the study; and poor adherence to the study protocol. T2DM was defined in accordance with the 2010 American Diabetes Association criteria [8].

This study was approved by the Ethics Committees of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. This trial was registered at www.chictr.org.cn with clinical trial registration number ChiCTR1800016786. All subjects provided written informed consent before enrollment.

2.2 Study protocol

This study was an open, randomized, intervention clinical trial that was conducted at the Department of Endocrinology and Metabolism of Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. Subjects willing to participate in the study were invited to undergo a 75 g OGTT, MRI, and MRS during study enrollment (visit 1, 3 days prior to commencement of interventions) and complete standardized questionnaires, including basic information, medical history, daily diet, and exercise habits. Anthropometric parameters and venous blood samples were collected. Acceptability and adherence to the study procedures were also evaluated by an experienced investigator. Those who met the eligibility criteria were randomly divided by another independent investigator into an LCD group (LCD) and an exercise group (EX) via the random number table method with an allocation ratio of 1:1.

The study team enrolled the participants and assigned them to interventions. The first glucose monitor was put on to continuously record glucose levels for 14 days after finishing MRI. All subjects were instructed to maintain their usual living habits after visit 1, and then the intervention was started 3 days later by adopting LCD or performing exercise for 3 weeks. Before the interventions, an experienced investigator assessed each subject’s medical history and clinical characteristics to ensure his or her eligible physical status and ability to finish the study. The following visits were conducted at the end of 1st week (visit 2, day 8), 2nd week (visit 3, day 15), and 3rd week (visit 4, day 22) of the intervention. The glucose monitor was replaced by a new one during visits 3 and 4 according to its valid time and the subjects’ available time. Therefore, two sensors were applied in each subject during entire trial period, and glucose values were recorded before, during, and after the interventions. The subjects were required to return to their former usual lifestyle for another 3 days after visit 4 and attend the last visit (visit 5). Endpoint MRI and MRS were conducted during visit 5 because CGM sensors would be affected during the examinations. The schematic study design is illustrated in Fig. 1.

First, the subjects in the LCD group were required to limit carbohydrate intake to<50 g/day with no restriction of calories, no restriction of proportion of fat and protein or water intake. The intake of fruits and various processed foods with a high sugar content or a high glycemic index were not allowed. Moreover, they were instructed to record their daily diet by taking photographs and sending those to the investigators before each meal. They were also asked to fill in record forms about their meal time and type and amount of food consumed, which were checked during the follow-up visits. Finally, they were told to not perform additional exercise.

With regard to the subjects in the EX group, they were required to not change their usual diet and refer to Chinese dietary guidelines [9]. Afterward, they were asked to perform moderate-to-vigorous exercise for no less than 30 min per day maintaining a heart rate of 60%–70% of their maximum predicted heart rate, which was calculated as 220 (210 for females) minus the subject’s age per minute [10]. Subsequently, they were instructed to perform aerobic exercises and supplement them with anaerobic exercises, including, but not limited to, jogging, trotting, skipping, cycling, and resistance exercise. Finally, they were obliged to record their daily exercise by taking photographs while exercising or sending records via mobile applications to the investigators every day. They were asked to fill in record forms about their exercise types, time, and duration. The forms were checked during the follow-up visits.

2.3 Body composition measurement

Bioelectrical impedance was assessed using Tanita TBF-418 body composition analyzer (Tanita Corporation, Tokyo, Japan). Fat percentage, fat mass (FM), fat free mass (FFM) of the whole body and segmental parts (trunk, right arm, left arm, right leg, and left leg), and total body water (TBW) were measured using the standard settings after manually imputing the measured height, sex, age, and somatotype of the subject (standard). Prior to the measurements, the subjects were instructed to fast overnight. During the measurements, the subjects were asked to wear minimal clothing and instructed to stand still barefoot, with their feet precisely touching the metal plates.

2.4 MRI and MRS

Liver steatosis and body fat content, including visceral and subcutaneous adipose tissues, were measured using a 3.0T clinical MRI scanner (Archiva, Philips Medical System, Amsterdam, the Netherlands) via the segmented breath–hold method. The scans were performed by an experienced radiologist who was blinded to the intervention groups. Sagittal, coronal, and axial slices were acquired.

MRI scans were conducted at the abdominal level between the L4 and L5 vertebrae in the supine position by a T1-weighted spin echo sequence with fluid-attenuated inversion recovery for MRI, where fat shows high signal intensity. Scan parameters were as follows: slice thickness, 1 cm; field of view, 42 cm × 42 cm; layer numbers, 6. Each scan lasted 10 s. A single image slice marked by the umbilicus was chosen for image segmentation. Image segmentation was performed twice by two trained investigators by using the SliceOmatic image analysis software (version 4.2, Tomovision Inc., Montreal, QC, Canada) to determine visceral fat area (VFA) and subcutaneous fat area (SFA). When the results differed by>10%, a third observer who did not know the results reanalyzed the images, and the mean of the two closer values was calculated as the final result.

The right posterior lobe of liver was selected in the MRS scans by using point-resolved single-voxel spectroscopy pulse sequence without water suppression with the following parameters: repetition time, 2000 ms; echo time, 50 ms; region of interest, 20 mm × 20 mm [11,12]. Each scan lasted 12 s. Spectrum processing was also performed twice by two trained investigators by using the standard Philips curve-fitting package to derive average intrahepatic triglyceride content (IHTC) (Fig. S1). The average IHTC was calculated as the signal from fat protons divided by the combined signal from fat and water protons in the liver. Subjects with IHTC≥5.56% were diagnosed with nonalcoholic fatty liver disease (NAFLD) [13].

2.5 CGM

Glucose levels of interstitial fluid were continuously monitored for up to 14 days and glucose data were automatically stored every 15 min by using Free Style Libre Pro Flash professional (Abbott®). The sensors were attached to the back of the left upper arm of each subject, and the values were blinded to them. Various parameters, including 24 h mean sensor glucose (MSG), daytime MSG (6:00–22:00), nighttime MSG (22:00–6:00), coefficient of variability (CV), and largest amplitude of glycemic excursions (LAGE), were calculated from the CGM data. CV was calculated from the ratio of standard deviation of glucose to MSG, and LAGE was the difference between the maximum and the minimum blood glucose concentrations within a day.

2.6 Anthropometric and biochemical measurements

Anthropometric and biochemical parameters were measured during the five clinical follow-up visits with the subjects fasting overnight (>10 h). Waist circumference was measured midway between the lowest border of the rib cage and the upper border of the iliac crest at the end of a normal expiration. Hip circumference was measured at the level of the largest circumference between the waist and thighs. Waist-to-hip ratio (WHR) was calculated. BMI was calculated as the weight (kg) divided by squared height (m).

Venous blood samples were collected and tested by laboratory technicians who were blinded to the intervention groups. FPG and 2hPG were assayed via the standard glucose oxidase method. Glycated albumin (GA), liver function, renal function, and serum lipids were measured by applying standard enzymatic methods by using a biochemical analyzer (Hitachi 7600, Tokyo, Japan). HbA_1c was measured via high-performance liquid chromatography with a VARIANT II Hemoglobin A_1c analyzer (Bio-Rad Laboratories, Hercules, CA). Serum insulin was measured via electrochemiluminescence immunoassay on a Cobas e601 analyzer (Roche Diagnostics GmbH, Mannheim, Germany). Insulin sensitivity was evaluated via homeostasis model assessment−insulin resistance (HOMA-IR) by using the following formula: HOMA-IR= FPG (mmol/L) × fasting serum insulin (FINS, μU/mL) / 22.5.

2.7 Statistical analyses

In this study, the primary outcome was change in body weight. Power analysis was conducted by using a two-sided test and an α level of 0.05; 22 subjects per group were required to provide 90% power to detect differences between the groups based on a dropout rate of 20%. The proposed mean±SD of change in body weight (LCD, 4.90±0.65 kg; EX, 4.33±0.32 kg) were based on data from previous studies [3,4]. Statistical analyses were performed using SPSS software version 20.0 (SPSS Inc., Chicago, IL, USA). The final results are presented as the mean±SEM unless otherwise specified. Normal distribution was checked by Kolmogorov–Smirnov test. Continuous variables at baseline were compared between the two groups by independent-samples t test. Comparison of categorical variables between the groups was performed by χ² tests. Differences in the parameters before and after the interventions were analyzed by paired t test. A P value (two sides)<0.05 was considered statistically significant.

3 Results

3.1 Characteristics of the study subjects

A total of 45 subjects met the eligibility criteria and were randomly divided into two groups (LCD, n = 22; EX, n = 23). All subjects in the LCD group followed the instructions and completed the study. One subject in the EX group failed to finish the intervention program because of issues at work. Finally, 22 participants (100%, 13 males) and 22 participants (95.7%, 16 males) completed the study in the LCD and EX groups, respectively (Fig. 2). No difference in gender was observed, and all the baseline characteristics were not substantially different between the two groups (Table 1).

Total daily calories of the LCD group were 1858.5±53.4 kcal in the first 3 days, and it declined to 1523.9±32.8 kcal during the LCD intervention. The baseline macronutrients in the LCD group were carbohydrates (48.5%±2.7% of total energy), proteins (21.1%±1.2% of total energy), and fats (30.4%±1.9% of total energy). The macronutrient compositions of carbohydrates, proteins, and fats changed to 11.4%±1.6%, 32.6%±1.7%, and 56.0%±2.3%, respectively, during the LCD intervention (Fig. S2).

3.2 Body composition, VFA, and SFA

After the 3-week intervention, body weight, BMI, total fat (%), and total FM significantly decreased in both groups (all P<0.05, Tables 1 and S1). The subjects in the LCD group lost 2.32 kg more weight (P<0.001) and 0.85 kg more FM (P<0.001) than those in the EX group. The body weight in the LCD group quickly decreased by 2.40±0.94 kg just after the 1st week of intervention and continued to fall in the following 2 weeks. However, the effect was weak in the EX group (Fig. 3A). In addition, the total FFM and TBW in the LCD group were reduced compared with baseline (Table S1). In this study, anhydrous lean body mass (ALBM) was equal to total FFM minus TBW. ALBM in the LCD group did not change, but it was lower in the EX group than its baseline after the intervention. No significant changes in total FFM and TBW were observed in the EX group. The comparison of changes in body composition between the two groups is given in Fig. 3B−3D.

Both groups had significant decreases in waist circumference, WHR, VFA, and SFA (all P<0.05, Table 1). However, the reduction in waist circumference and WHR in the LCD group was greater than that in the EX group. No significant differences in the reduction in SFA or VFA were observed between the two groups.

Three days after stopping the LCD intervention (visit 5), body weight increased to 80.9±3.03 kg, which was significantly higher than the endpoint (P = 0.001). Waist circumference and WHR also increased. However, they did not change 3 days after the subjects stopped exercising (Table S2).

3.3 Glucose and lipid metabolism and renal function

During the intervention, the FPG in both groups remained relatively stable. Only on day 8 did the FPG of the LCD group decrease compared with baseline (day 8, 4.81±0.08 mmol/L vs. day 0, 5.04±0.09 mmol/L; P = 0.017) and was also lower than that of the EX group (Fig. 4A). FINS and HOMA-IR sharply declined after 1-week LCD intervention (FINS: day 8, 8.20±1.00 μU/mL vs. day 0, 13.88±1.36 μU/mL, P<0.001; HOMA-IR: day 8, 1.80±0.24 vs. day 0, 3.13±0.32, P<0.001), reaching the trough of wave and had a slight rising trend in the following 2 weeks, although they were still lower than their baselines (Fig. 4B and 4C). FINS and HOMA-IR also significantly decreased after 1-week of exercise. However, they continued to gradually decline in the 2nd and 3rd weeks. In addition, GA, which indicates glycemic control in the past 2–3 weeks, decreased in the LCD group (P = 0.023, Table 1). However, 3 days after the intervention, FPG, FINS, and HOMA-IR in the LCD group immediately rose to their baseline levels, whereas they remained decreased in the EX group. Spearman correlation analysis also revealed no significant correlation between changes in BMI and body composition measurements with changes in insulin resistance in both groups (Table S3).

Triglycerides decreased in both groups (LCD, P = 0.015; EX, P = 0.037; Table 1). Total cholesterol only decreased in the EX group (P = 0.021). High-density lipoprotein cholesterol levels did not significantly change in the two groups, although the levels showed an upward tendency. Low-density lipoprotein cholesterol levels increased in the LCD group (P = 0.019) and decreased in the EX group (P = 0.040).

In terms of renal function, blood urea nitrogen (BUN) and serum creatinine (SCr) were elevated in the LCD group. Among electrolytes, only serum calcium (Ca) in the LCD group significantly increased. Serum potassium (K), serum sodium (Na), and serum chloride (Cl) in both groups remained constant (Table S4).

3.4 Liver function and hepatic steatosis

Serum alanine aminotransferase (ALT) and γ-glutamyl transferase (γ-GT) decreased in both groups (all P<0.05, Table 2), whereas serum alkaline phosphatase (AKP) levels only declined in the LCD group (P = 0.002). The ratio of albumin and globulin (ALB/GLB) increased and pre-albumin (PALB) decreased in both groups. However, total protein level and ALB level in the LCD group was higher than those in the EX group.

After the intervention, the IHTC of all subjects in both groups significantly decreased (LCD, P<0.001; EX, P = 0.019; Table 2) and decreased by 3.23%±0.78% in the LCD group and 3.31%±1.30% in the EX group. No differences were observed between the two groups. According to the diagnostic criteria of hepatic steatosis, 12 subjects in the LCD group and 8 subjects in the EX group were diagnosed with NAFLD. Further analysis revealed that the IHTC in those with NAFLD all declined (Fig. 5A) in endpoint. Among them, the liver fat content of five subjects in the LCD group (41.7%) and three subjects in the EX group (37.5%) returned to normal range. The comparison of IHTC in these 20 subjects is shown in Fig. 5B.

3.5 Subgroup analysis of gender and BMI

The influence of gender and BMI on the intervention results was further analyzed by grouping the subjects according to these parameters (24 kg/m²≤BMI<28 kg/m² or BMI≥28 kg/m²). In the subgroup of gender, baseline body weight and waist circumference of males in both groups were larger than those of females (Tables S5 and S6). However, no gender difference was observed in the change of anthropometric parameters. Glucose and lipid metabolism among males seemed to improve more, but most of them had no statistical difference. In addition, according to their baseline BMI, 13 subjects in the LCD group and 16 in the EX group were overweight. Moreover, nine subjects in the LCD group and six in the EX group were obese. Results showed that only the body weight of the subjects with obesity decreased more, and no significant difference was found in the other parameters (Tables S7 and S8).

3.6 CGM and glycemic control

A total of 44 subjects were monitored in this study. The data of 12 subjects were incomplete because of sensor malfunction, accidental strike to sensors, and other unknown reasons. Finally, the CGM data of 32 subjects, including 16 in the LCD group and 16 in the EX group, were analyzed. In addition, sensor data on days 11, 12, 13, and 14 were not included, during which the sensors were replaced and the data were incomplete and inaccurate.

MSG levels of the two groups before, during, and after the interventions are illustrated in Fig. 6A. Baseline MSG levels did not differ between the two groups. Once LCD intervention was started, MSG level significantly declined on day 1 (day 0, 4.50±0.11 mmol/L; day 1, 4.14±0.10 mmol/L; P<0.001) and was lower than that of the EX group (day 1, 4.85±0.11 mmol/L; P = 0.008, Fig. 6B). The average MSG level also decreased during the intervention. However, once the LCD intervention was stopped, MSG level quickly rebounded to baseline level. The MSG level of the EX group remained the same during the entire study. Daytime and nighttime MSG levels on day 1 and during the intervention also decreased compared with baseline level and were lower than those in the EX group (all P<0.01, Fig. 6C).

The 24 h glucose fluctuations of the two groups before and during the intervention are plotted in Fig. S3a and S3b. The blue curve of the LCD group was below its baseline curve and considerably more stable. By contrast, the curves of the EX group and its baseline closely overlapped, although the fluctuations of the EX group was slightly lesser than its baseline curve. Glucose variability was further analyzed by calculating the CV during the entire study period from the CGM data (Fig. S3c). No difference between baseline CVs in the two groups was observed (LCD, 17.89%±1.25%; EX, 21.77%±1.87%; P = 0.091). It sharply fell to 10.10%±1.00% on day 1 in the LCD group and was significantly lower than that of the EX group (19.55%±1.26%, P<0.001). Although the average CV during the intervention in the EX group was also higher than that in the LCD group (EX, 18.43%±0.36%; LCD, 12.14%±0.32%; P<0.001), it was significantly lower than its baseline (P= 0.035). The CV on day 22 rose to baseline after the subjects in the LCD group returned to their usual diets (day 0, 17.89%±1.25%; day 22, 21.21%±1.93%; P = 0.099). Furthermore, LAGE in the LCD group on day 1 and during the intervention also significantly decreased compared with baseline (baseline, 3.69±0.30 mmol/L vs. day 1, 2.03±0.18 mmol/L or average, 2.36±0.07 mmol/L; both P<0.001), but it did not change in the EX group (Fig. S3d).

4 Discussion

This study was the first to compare the effects of LCD and exercise interventions on obesity. Results suggested that LCD might have a stronger effect on weight loss than exercise. Results revealed that the larger part of weight loss after LCD interventions was from fat mass. Several studies [14–17] also assessed the effects of LCDs on obesity. Moreno et al. [15] reported that 88% of the participants with obesity in the very low-calorie-ketogenic diet group lost over 10% of their initial weight, and their lean mass was unaffected. However, some studies [18–20] showed a reduction in both body fat and lean mass associated with LCD. The discrepancy of the results might be attributed to different metabolic characteristics of the participants, methods of body composition measurements, and duration of studies. Another important point that must be discussed is energy intake. Total daily calories declined during the LCD intervention, whereas energy intake in the EX group was supposed to remain relatively stable. The reduction in energy was possibly one of the major contributors to the difference in weight loss recorded between the two groups. In addition, loss of trunk fat was relatively similar in the two groups, whereas exercise caused lesser loss of limb fat. Our results agreed with those of Ramírez-Campillo et al. [21], who showed a reduction in trunk and arm fat mass with no substantial reduction in fat mass of lower extremities after 12 weeks of localized leg training. The findings of previous findings were consistent with these results [22–24]. No considerable fat loss of limbs was observed because of muscle hypertrophy for the trained extremities, whereas adipose cell size did not decrease [24].

The reason the weight loss effect of LCD intervention was more remarkable than exercise may be ascribed to several factors. Fluid–electrolyte metabolism is probably one of the major causes. First, the decreased consumption of carbohydrates and the increased consumption of fats and proteins may cause natriuresis, kaliuresis, and diuresis [25–27]. Moreover, severe restriction of carbohydrate intake may deplete glycogen storage and result in the excretion of bound water [28]. These mechanisms perhaps explain why body water sharply decreased after LCD intervention but remained stable during the exercise intervention. Moreover, the decrease in serum insulin concentration might be another reason because it can promote lipolysis, resulting in reduced fat storage [26]. Previous studies also suggested that less proportion of carbohydrates in the diet can promote energy expenditure [29–31]. In addition, LCD consumption is reportedly related to enhanced satiety and reduced peptide, both of which would further decrease energy intake, thereby intensifying energy deficit [32–34].

Both LCDs and exercises can reportedly alleviate hepatic steatosis in patients with NAFLD [4,35–38]. However, no head-to-head study has compared the impacts of the two methods. The present results showed that liver fat content was reduced by 27.1% by LCD, and the IHTC in 41.7% of the subjects with NALFD returned to normal range after the intervention. Hepatic enzymes, including ALT, AKP, and γ-GT, decreased. After exercise, IHTC was alleviated by 30.6%, and the number of subjects with NALFD decreased by 37.5% as hepatic enzymes decreased, indicating that LCD and exercise had similar effects on improving liver metabolism. However, this result was different from the stronger effect of LCD on weight loss mentioned above. Given that CGM sensors are vulnerable to magnetic resonance, the endpoint MRS was not conducted right after the intervention and was instead postponed after 3 days of habitual diet when the CGM sensors had been removed. With regard to serum protein levels, elevated ALB after the LCD intervention may be due to the increase in proportions of protein consumption in daily meals. The increases in BUN and SCr could be explained by the same reason. However, PALB fell after both interventions, and the underlying mechanism might be related to unstable nutritional status, such as impairment of energy intake, weight loss, loss of muscle mass, and loss of subcutaneous fat [39].

Previous studies demonstrated that LCD can reduce HbA_1c levels and improve glycemic fluctuations in individuals with T2DM [40–42]. However, few studies have applied CGM to investigate the impact of LCDs on nondiabetic people, let alone covered the entire intervention time. The subjects in this trial were continuously monitoring for nearly 4 weeks. CGM can not only reflect overall glucose profile but is also more highly sensitive in detecting glucose fluctuations compared with traditional methods. Results showed that all the parameters that reflected glycemic control, including MSG, CV, LAGE, and GA, decreased after LCD intervention, indicating its superior role in improving glucose homeostasis compared with exercise. Moreover, the changes in glucose and its variability in subjects consuming LCD were quick to act, which were characterized by sharp decline on the 1st day and a rapid rise right after having normal diets. Samkani et al. [43] showed that responding to a carbohydrate-reduced high-protein meal, peak glucose concentrations, glucose excursions, and insulin secretion rate substantially decreases after ingestion compared with a high carbohydrate meal. Both of these studies indicated that carbohydrates are directly involved in glucose metabolism. Therefore, restriction or intake of staple food, such as rice, will result in rapid glucose excursion. The same can be said of insulin and HOMA-IR. Low or high consumption of carbohydrates leads to a corresponding reduction or induction in insulin secretion, whereas the impact of exercise is relatively gentle and stable.

Few studies continuously monitored glucose levels during regular exercise in individuals with obesity. They used CGM to investigate the instant reaction of blood glucose to a certain intensity of exercise regimen that lasted for only several hours [44,45]. In the present study, CGM was performed for 21 consecutive days of exercise. Although MSGs did not decrease and were higher than taking LCDs during exercise, glucose fluctuations slightly improved, suggesting the long-term regular exercise has a beneficial role in glycemic control.

Notably, the strong effect of LCDs on weight loss and glucose metabolism quickly degenerated just after 3 days of normal diet, which was characterized by an increase in body weight and an increase in glycemic parameters, including FINS, HOMA-IR, MSG, and CV. This phenomenon suggested that LCD exerted its protective effects on metabolism only during the diet period. Its effectiveness began to lessen once the LCD intervention was stopped. By contrast, the legacy effect of exercise was remarkable. Body weight and glucose metabolism remained stable in spite of the termination of exercise. This study revealed the poor legacy effect of LCD. The issue of whether long-term LCD intervention could sustain its beneficial effects after the intervention warrants further investigation.

This study has several limitations. First, the sample size was relatively small. Therefore, further large-scale randomized clinical trials with longer follow-up durations are needed to explore the effects of LCDs and exercise on obesity. Moreover, this study lack mechanistic investigation into the possible underlying mechanisms of the effect of LCD on weight loss. Finally, the data of liver fat were collected from a small portion of the liver, which might have led to a sampling error [46].

In conclusion, this pilot study suggested that both 3-week LCD and exercise interventions could alleviate liver steatosis and insulin resistance. LCDs might be more efficient than exercise in terms of reducing weight and regulating glucose homeostasis.

References

[1] The GBD 2015 Obesity Collaborators. Health effects of overweight and obesity in 195 countries over 25 years. N Engl J Med 2017; 377(1): 13–27

[2] Nordmann AJ, Nordmann A, Briel M, Keller U, Yancy WS Jr, Brehm BJ, Bucher HC. Effects of low-carbohydrate vs low-fat diets on weight loss and cardiovascular risk factors: a meta-analysis of randomized controlled trials. Arch Intern Med 2006; 166(3): 285–293

[3] Bazzano LA, Hu T, Reynolds K, Yao L, Bunol C, Liu Y, Chen CS, Klag MJ, Whelton PK, He J. Effects of low-carbohydrate and low-fat diets: a randomized trial. Ann Intern Med 2014; 161(5): 309–318

[4] Zhang HJ, He J, Pan LL, Ma ZM, Han CK, Chen CS, Chen Z, Han HW, Chen S, Sun Q, Zhang JF, Li ZB, Yang SY, Li XJ, Li XY. Effects of moderate and vigorous exercise on nonalcoholic fatty liver disease: a randomized clinical trial. JAMA Intern Med 2016; 176(8): 1074–1082

[5] Chen HT, Chung YC, Chen YJ, Ho SY, Wu HJ. Effects of different types of exercise on body composition, muscle strength, and IGF-1 in the elderly with sarcopenic obesity. J Am Geriatr Soc 2017; 65(4): 827–832

[6] Carlson AL, Mullen DM, Bergenstal RM. Clinical use of continuous glucose monitoring in adults with type 2 diabetes. Diabetes Technol Ther 2017; 19(S2): S4–S11

[7] Jia W, Weng J, Zhu D, Ji L, Lu J, Zhou Z, Zou D, Guo L, Ji Q, Chen L, Chen L, Dou J, Guo X, Kuang H, Li L, Li Q, Li X, Liu J, Ran X, Shi L, Song G, Xiao X, Yang L, Zhao Z; Chinese Diabetes Society. Standards of medical care for type 2 diabetes in China 2019. Diabetes Metab Res Rev 2019; 35(6): e3158

[8] American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010; 33(Suppl 1): S62–S69

[9] Chinese Nutrition Society. Chinese Dietary Guidelines Summary. 1st ed. Beijing: People’s Medical Publishing Press, 2016

[10] Hill J, Timmis A. Exercise tolerance testing. BMJ 2002; 324(7345): 1084–1087

[11] Reeder SB, Cruite I, Hamilton G, Sirlin CB. Quantitative assessment of liver fat with magnetic resonance imaging and spectroscopy. J Magn Reson Imaging 2011; 34(4): 729–749

[12] Qian LL, Wu L, Zhang L, Zhang J, Zhou J, Li YH, Fang QC, Li HT, Jia WP. Serum biomarkers combined with ultrasonography for early diagnosis of non-alcoholic fatty liver disease confirmed by magnetic resonance spectroscopy. Acta Pharmacol Sin 2020; 41(4): 554–560

[13] Szczepaniak LS, Nurenberg P, Leonard D, Browning JD, Reingold JS, Grundy S, Hobbs HH, Dobbins RL. Magnetic resonance spectroscopy to measure hepatic triglyceride content: prevalence of hepatic steatosis in the general population. Am J Physiol Endocrinol Metab 2005; 288(2): E462–E468

[14] Merra G, Miranda R, Barrucco S, Gualtieri P, Mazza M, Moriconi E, Marchetti M, Chang TF, De Lorenzo A, Di Renzo L. Very-low-calorie ketogenic diet with aminoacid supplement versus very low restricted-calorie diet for preserving muscle mass during weight loss: a pilot double-blind study. Eur Rev Med Pharmacol Sci 2016; 20(12): 2613–2621

[15] Moreno B, Bellido D, Sajoux I, Goday A, Saavedra D, Crujeiras AB, Casanueva FF. Comparison of a very low-calorie-ketogenic diet with a standard low-calorie diet in the treatment of obesity. Endocrine 2014; 47(3): 793–805

[16] Hashimoto Y, Fukuda T, Oyabu C, Tanaka M, Asano M, Yamazaki M, Fukui M. Impact of low-carbohydrate diet on body composition: meta-analysis of randomized controlled studies. Obes Rev 2016; 17(6): 499–509

[17] Shai I, Schwarzfuchs D, Henkin Y, Shahar DR, Witkow S, Greenberg I, Golan R, Fraser D, Bolotin A, Vardi H, Tangi-Rozental O, Zuk-Ramot R, Sarusi B, Brickner D, Schwartz Z, Sheiner E, Marko R, Katorza E, Thiery J, Fiedler GM, Blüher M, Stumvoll M, Stampfer MJ; the Dietary Intervention Randomized Controlled Trial (DIRECT) Group. Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. N Engl J Med 2008; 359(3): 229–241

[18] Noakes M, Foster PR, Keogh JB, James AP, Mamo JC, Clifton PM. Comparison of isocaloric very low carbohydrate/high saturated fat and high carbohydrate/low saturated fat diets on body composition and cardiovascular risk. Nutr Metab (Lond) 2006; 3(1): 7

[19] Brinkworth GD, Noakes M, Clifton PM, Buckley JD. Effects of a low carbohydrate weight loss diet on exercise capacity and tolerance in obese subjects. Obesity (Silver Spring) 2009; 17(10): 1916–1923

[20] Ruth MR, Port AM, Shah M, Bourland AC, Istfan NW, Nelson KP, Gokce N, Apovian CM. Consuming a hypocaloric high fat low carbohydrate diet for 12 weeks lowers C-reactive protein, and raises serum adiponectin and high density lipoprotein-cholesterol in obese subjects. Metabolism 2013; 62(12): 1779–1787

[21] Ramírez-Campillo R, Andrade DC, Campos-Jara C, Henríquez-Olguín C, Alvarez-Lepín C, Izquierdo M. Regional fat changes induced by localized muscle endurance resistance training. J Strength Cond Res 2013; 27(8): 2219–2224

[22] Gwinup G, Chelvam R, Steinberg T. Thickness of subcutaneous fat and activity of underlying muscles. Ann Intern Med 1971; 74(3): 408–411

[23] Kostek MA, Pescatello LS, Seip RL, Angelopoulos TJ, Clarkson PM, Gordon PM, Moyna NM, Visich PS, Zoeller RF, Thompson PD, Hoffman EP, Price TB. Subcutaneous fat alterations resulting from an upper-body resistance training program. Med Sci Sports Exerc 2007; 39(7): 1177–1185

[24] Krotkiewski M, Aniansson A, Grimby G, Björntorp P, Sjöström L. The effect of unilateral isokinetic strength training on local adipose and muscle tissue morphology, thickness, and enzymes. Eur J Appl Physiol Occup Physiol 1979; 42(4): 271–281

[25] Seshadri P, Iqbal N. Low carbohydrate diets for weight loss: historical & environmental perspective. Indian J Med Res 2006; 123(6): 739–747

[26] Rabast U, Vornberger KH, Ehl M. Loss of weight, sodium and water in obese persons consuming a high- or low-carbohydrate diet. Ann Nutr Metab 1981; 25(6): 341–349

[27] Azar GJ, Bloom WL. Similarities of carbohydrate deficiency and fasting. II. Ketones, nonesterified fatty acids and nitrogen excretion. Arch Intern Med 1963; 112(3): 338–343

[28] Astrup A, Meinert Larsen T, Harper A. Atkins and other low-carbohydrate diets: hoax or an effective tool for weight loss? Lancet 2004; 364(9437): 897–899

[29] Ebbeling CB, Swain JF, Feldman HA, Wong WW, Hachey DL, Garcia-Lago E, Ludwig DS. Effects of dietary composition on energy expenditure during weight-loss maintenance. JAMA 2012; 307(24): 2627–2634

[30] Ebbeling CB, Feldman HA, Klein GL, Wong JMW, Bielak L, Steltz SK, Luoto PK, Wolfe RR, Wong WW, Ludwig DS. Effects of a low carbohydrate diet on energy expenditure during weight loss maintenance: randomized trial. BMJ 2018; 363: k4583

[31] Hall KD, Chen KY, Guo J, Lam YY, Leibel RL, Mayer LE, Reitman ML, Rosenbaum M, Smith SR, Walsh BT, Ravussin E. Energy expenditure and body composition changes after an isocaloric ketogenic diet in overweight and obese men. Am J Clin Nutr 2016; 104(2): 324–333

[32] Hu T, Yao L, Reynolds K, Niu T, Li S, Whelton P, He J, Bazzano L. The effects of a low-carbohydrate diet on appetite: a randomized controlled trial. Nutr Metab Cardiovasc Dis 2016; 26(6): 476–488

[33] Gibson AA, Seimon RV, Lee CM, Ayre J, Franklin J, Markovic TP, Caterson ID, Sainsbury A. Do ketogenic diets really suppress appetite? A systematic review and meta-analysis. Obes Rev 2015; 16(1): 64–76

[34] Johnstone AM, Horgan GW, Murison SD, Bremner DM, Lobley GE. Effects of a high-protein ketogenic diet on hunger, appetite, and weight loss in obese men feeding ad libitum. Am J Clin Nutr 2008; 87(1): 44–55

[35] Mardinoglu A, Wu H, Bjornson E, Zhang C, Hakkarainen A, Räsänen SM, Lee S, Mancina RM, Bergentall M, Pietiläinen KH, Söderlund S, Matikainen N, Ståhlman M, Bergh PO, Adiels M, Piening BD, Granér M, Lundbom N, Williams KJ, Romeo S, Nielsen J, Snyder M, Uhlén M, Bergström G, Perkins R, Marschall HU, Bäckhed F, Taskinen MR, Borén J. An integrated understanding of the rapid metabolic benefits of a carbohydrate-restricted diet on hepatic steatosis in humans. Cell Metab 2018; 27(3): 559–571.e5

[36] Schwimmer JB, Ugalde-Nicalo P, Welsh JA, Angeles JE, Cordero M, Harlow KE, Alazraki A, Durelle J, Knight-Scott J, Newton KP, Cleeton R, Knott C, Konomi J, Middleton MS, Travers C, Sirlin CB, Hernandez A, Sekkarie A, McCracken C, Vos MB. Effect of a low free sugar diet vs usual diet on nonalcoholic fatty liver disease in adolescent boys: a randomized clinical trial. JAMA 2019; 321(3): 256–265

[37] Jang EC, Jun DW, Lee SM, Cho YK, Ahn SB. Comparison of efficacy of low-carbohydrate and low-fat diet education programs in non-alcoholic fatty liver disease: a randomized controlled study. Hepatol Res 2018; 48(3): E22–E29

[38] Hashida R, Kawaguchi T, Bekki M, Omoto M, Matsuse H, Nago T, Takano Y, Ueno T, Koga H, George J, Shiba N, Torimura T. Aerobic vs. resistance exercise in non-alcoholic fatty liver disease: a systematic review. J Hepatol 2017; 66(1): 142–152

[39] Smith SH. Using albumin and prealbumin to assess nutritional status. Nursing 2017; 47(4): 65–66

[40] Tay J, Luscombe-Marsh ND, Thompson CH, Noakes M, Buckley JD, Wittert GA, Yancy WS Jr, Brinkworth GD. A very low-carbohydrate, low-saturated fat diet for type 2 diabetes management: a randomized trial. Diabetes Care 2014; 37(11): 2909–2918

[41] Blaychfeld-Magnazi M, Reshef N, Zornitzki T, Madar Z, Knobler H. The effect of a low-carbohydrate high-fat diet and ethnicity on daily glucose profile in type 2 diabetes determined by continuous glucose monitoring. Eur J Nutr 2020; 59(5): 1929–1936

[42] Tay J, Luscombe-Marsh ND, Thompson CH, Noakes M, Buckley JD, Wittert GA, Yancy WS Jr, Brinkworth GD. Comparison of low- and high-carbohydrate diets for type 2 diabetes management: a randomized trial. Am J Clin Nutr 2015; 102(4): 780–790

[43] Samkani A, Skytte MJ, Thomsen MN, Astrup A, Deacon CF, Holst JJ, Madsbad S, Rehfeld JF, Krarup T, Haugaard SB. Acute effects of dietary carbohydrate restriction on glycemia, lipemia and appetite regulating hormones in normal-weight to obese subjects. Nutrients 2018; 10(9): 1285

[44] Hatamoto Y, Goya R, Yamada Y, Yoshimura E, Nishimura S, Higaki Y, Tanaka H. Effect of exercise timing on elevated postprandial glucose levels. J Appl Physiol (1985) 2017; 123(2): 278–284

[45] Sylow L, Kleinert M, Richter EA, Jensen TE. Exercise-stimulated glucose uptake- regulation and implications for glycaemic control. Nat Rev Endocrinol 2017; 13(3): 133–148

[46] Wong VW, Adams LA, de Lédinghen V, Wong GL, Sookoian S. Noninvasive biomarkers in NAFLD and NASH—current progress and future promise. Nat Rev Gastroenterol Hepatol 2018; 15(8): 461–478

RIGHTS & PERMISSIONS

Higher Education Press