Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2018 Aug 13.
Published in final edited form as: J Proteome Res. 2013 Nov 18;12(12):5801–5811. doi: 10.1021/pr4008199

Very Low Carbohydrate Diet Significantly Alters the Serum Metabolic Profiles in Obese Subjects

Yunjuan Gu †, Aihua Zhao †, Fengjie Huang ǁ, Yinan Zhang †, Jiajian Liu †, Congrong Wang †, Wei Jia †,‡, Guoxiang Xie †,‡,*, Weiping Jia †,*
PMCID: PMC6088239  NIHMSID: NIHMS982422  PMID: 24224694

Abstract

Emerging evidence has consistently shown that very low carbohydrate diet (VLCD) can protect against the development of obesity, but the underlying mechanisms are not fully understood. Here we applied a comprehensive metabonomics approach using an ultra-performance liquid chromatography-quadrupole time of flight mass spectrometry and gas chromatography-time of flight mass spectrometry to study the effects of an 8-week dietary intervention with VLCD on serum metabolic profiles in obese subjects. The VLCD intervention resulted in a weight loss and significantly decreased homeostasis model assessment-insulin resistance. The metabonomics analysis identified a number of differential serum metabolites (p < 0.05) primarily attributable to fatty acids, amino acids including branched chain amino acids, amines, lipids, carboxylic acids, and carbohydrates in obese subjects compared to healthy controls. The correlation analysis among time, VLCD intervention and clinical parameters revealed that the changes of metabolites correlated with the changes of clinical parameters and showed difference in males and females. Fatty acids, amino acids and carboxylic acids were increased in obese subjects compared with their normal healthy counterparts. Such increased levels of serum metabolites were attenuated after VLCD intake, suggesting that the health beneficial effects of VLCD are associated with attenuation of impaired fatty acid and amino acid metabolism. It also appears that VLCD induced significant metabolic alterations independent of the obesity-related metabolic changes. The altered metabolites in obese subjects post-VLCD intervention include arachidonate, cis-11,14-eicosadienoate, cis-11,14,17-eicosatrienoate, 2-aminobutyrate, acetyl-carnitine, and threonate, all of which are involved in inflammation and oxidation processes. The results revealed favorable shifts in fatty acids and amino acids after VLCD intake in obese subjects, which should be considered biomarkers for evaluating health beneficial effects of VLCD and similar dietary interventions.

Keywords: very low carbohydrate diet; obesity, metabonomics; ultra performance liquid chromatography-quadrupole time of flight mass spectrometry; gas chromatography-time of flight mass spectrometry

Table of Contents Synopsis

graphic file with name nihms-982422-f0001.jpg

Introduction

Obesity has reached epidemic proportions worldwide 1 and is strongly linked to the development of diabetes, hypertension, cardiovascular disease, coronary heart disease, stroke, and several types of cancer.2 Because of the severe comorbidities of obesity, people attempt to lose weight through several alternative diets. Recently, there has been a resurgence of interest in very low carbohydrate diet (VLCD) as a means of weight loss and metabolic improvements. Evidence from clinical studies and meta analyses suggested that VLCD can decrease body weight, improve metabolic parameters, insulin resistance/sensitivity and nonalcoholic fatty liver disease (NAFLD).3 Alternative diets (high protein, low-carbohydrate, high-fat) produce significantly greater weight loss in the short-term (6 months) compared to the conventional fat-restricted diet 4, 5 with no strict control of total energy intake. This significantly greater weight loss is likely due to spontaneous reduction in energy intake 5, 6 which may be linked to a lack of diet variety and changes in humoral satiety factors.7 However, when energy intake is strictly controlled and reduced to a hypocaloric level no difference in weight change is detectable between alternative and high-carbohydrate diets,8, 9 suggesting that primarily calorie restriction and not macronutrient composition is responsible for weight loss in hypo-caloric diets.9, 10 It was reported that a weight-maintaining, high-protein diet was associated with improvements in overall glucose control as postprandial blood-glucose concentrations and glycated haemoglobin decreased significantly compared to a conventional high-carbohydrate diet.11 In contrast, a study showed that glycated haemoglobin and fasting plasma glucose decreased and insulin sensitivity increased in the high-carbohydrate but not the high-protein group, while weight loss in both groups was comparable.8 We have shown that VLCD intervention induced rapid weight reduction with decreased total abdominal subcutaneous and visceral adipose tissue compartments, and liver fat content, increased skeletal muscle percentage of whole body weight, improved metabolic profile and insulin resistance and sensitivity.12 There is growing evidence that obesity and related conditions are characterized by a broad perturbation of metabolic physiology involving considerable changes in amino acid (branched chain amino acid (BCAA), and aromatic amino acids) and fatty acid metabolism 13–15 in addition to glucose.16 This new evidence is prompting the application of methods monitoring a broad range of molecular species, i.e. metabolomics, to study the beneficial effects of potentially health-promoting diets.15, 17

Metabonomics has been applied to investigate the effects of dietary carbohydrate modification on human serum metabolic profiles.18 The application of metabonomics to well-designed controlled intervention studies can be a useful tool to elucidate the complex physiological effects of VLCD, which might help in understanding their beneficial effects on human health.19, 20

Here we analyzed the serum metabolites in healthy controls and obesity subjects before and after VLCD intervention by ultra performance liquid chromatography quadrupole time of flight mass spectrometry (UPLC-QTOFMS) and gas chromatography-time of flight mass spectrometry (GC-TOFMS) to determine metabolic differences between obese and healthy human subjects and investigate the effect of VLCD intervention on serum metabolic profiles in obese subjects.

Materials and Methods

Study populations.

A number of 45 “healthy” obese (aged 17.8–52.0 years, mean BMI of 32.58 kg/m2) and 30 healthy control (aged 23.4–43.8 years, mean BMI of 21.29 kg/m2) subjects was recruited from the outpatient clinic of endocrinology and metabolism department of Shanghai Jiao Tong University affiliated Sixth People’s Hospital. The exclusion criteria were as follows12: pregnant or plan for pregnant; lactation or postmenopausal women; use of any prescription medication in previous 2 months; had any weight loss diet or pill during the past 6 months; consuming > 20 g/day of alcohol; tobacco use within 6 months; cardiovascular or endocrine disease history; hypertension history or current elevated blood pressure (systolic blood pressure [SBP]: ≥ 150 mm Hg; diastolic blood pressure [DBP] ≥ 90 mm Hg; current treatment for hypertension); diabetes mellitus; acute or chronic infections; liver disease, kidney disease, gastrointestinal disease or any other acute or chronic diseases requiring treatment.

The demographic information and clinical characteristics of all subjects are shown in Table 1. This study was approved by the Institutional Review Board of the Sixth People’s Hospital. All participants provided written informed consent.

Table 1.

Demographics and clinical characteristics of healthy controls and obese human subjects.

Characteristics Control Baseline Week 4 Week 8
n = 30 n = 45 n = 45 n = 38
Age (year) 28.21 ± 5.35 31.87 ± 8.98 a 31.87 ± 8.98 a 32.33 ± 9.30 a
Gender (Male/Female) 13/17 25/20 25/20 23/15
Height (cm) 165.33 ± 7.88 170.88 ± 8.69 a 170.83 ± 8.64 a 171.54 ± 8.82 a
Weight (kg) 58.37 ± 7.71 95.70 ± 18.67 a 89.83 ± 17.97 a 88.54 ± 18.01 a
BMI (kg/m2) 21.29 ± 1.75 32.58 ± 4.36 a 30.59 ± 4.21 a,b 29.88 ± 4.11 a,b
Waist (cm) 77.04 ± 6.86 104.24 ± 11.44 a 100.69 ± 10.58 a 99.21 ± 9.85 a,b
Hip (cm) 90.93 ± 3.93 110.56 ± 9.10 a 107.06 ± 8.93 a 106.84 ± 8.61 a
WHR 0.85 ± 0.06 0.94 ± 0.07 a 0.94 ± 0.05 a 0.93 ± 0.05 a
FPG (mmol/L) 4.76 ± 0.28 5.27 ± 0.88 a 4.99 ± 0.48 a,b 5.16 ± 0.42 a
2h PG (mmol/L) 5.54 ± 1.02 7.56 ± 1.73 a NA 6.87 ± 1.69
FINS (μU/mL) 7.06 ± 3.00 22.67 ± 27.54 a 10.25 ± 5.94 a,b 11.89 ± 8.93 a,b
2h INS (μU/mL) 50.53 ± 23.81 136.56 ± 75.87 a NA 71.88 ± 51.92 a,b
HOMA-IR 1.35 ± 0.77 6.16 ± 10.61 a 2.28 ± 1.48 a,b 2.74 ± 2.37 a,b
2h HOMA-IR 11.35 ± 7.55 49.19 ± 34.95 NA 21.51 ± 18.24
Open in a new tab
a,

p < 0.05, significantly different from healthy controls;

b,

p < 0.05, significantly different from baseline.

Experimental protocol.

One week before initiation of the study, all subjects were asked to maintain their habitual energy intake. At week 0, 4 and 8, serum samples were collected and anthropometric parameters, glucose concentration, and insulin resistance and sensitivity were measured. All study measurements were obtained before 10 A.M. after an overnight fast. Serum samples for metabonomics analysis were collected in the morning before breakfast and kept at −80 °C until analysis.

Dietary Intervention.

The obese subjects were subject to dietary intervention for two periods as reported in our previous study.12 In brief, energy intake was restricted to less than 800 Kcal/day (carbohydrate intake < 20 g/d). All daily meals were replaced as follows: a cup of soybean milk (200 ml) and a boiled egg at breakfast; a diet nutrition bar (106 Kcal: 2.8 g carbohydrate, 11.2 g protein and 5.6 g fat; Nutriease Health Technology Co, Ltd, Hangzhou, China), non-starchy vegetables (< 200 kcal) and 50 g protein from meat (i.e. beef, lean pork, skinned chicken, fish) at lunch and dinner. Supplementation of multivitamins and minerals were provided every day.

Subjects were also encouraged to drink at least 1.8 liters of water per day, and asked to maintain their habitual level of physical activity. Regular telephone contact to individual by nutritionists was provided for nutrient support.

Anthropometric Measurements.

Body weight and height were measured using standard methods for the calculation of BMI (kg/m2). Hip and waist circumferences were measured for the calculation of the waist/hip ratio (WHR).

Glucose, Insulin, and Homeostatic Model Assessment of Insulin Resistance (HOMA-IR).

Plasma glucose concentrations (fasting glucose, and 2 h postprandial plasma glucose) were measured by the glucose oxidase-peroxidase method using commercial kits (Shanghai Biological Products Institution, Shanghai, China) according to the manufacturer’s instructions. Serum insulin concentrations were measured using the radioimmunoassay method (Beijing North Institute of Biological Technology, Beijing, China). Insulin sensitivity was measured by HOMA, using the following formula: HOMA = (fasting insulin in mU/mL × fasting glucose in mM) / 22.5.

Serum Metabolomic Analysis.

Serum samples were prepared and analyzed by UPLC-QTOFMS and GC-TOFMS following our previously published protocols 21, 22 with minor modifications. Experimental details are provided in the Supporting Information.

Data Analysis and Statistics.

The acquired MS data from UPLC-QTOFMS and GC-TOFMS were analyzed according to our previously published work. The ESI positive and negative raw data generated from UPLC-QTOFMS was analyzed by the MarkerLynx applications manager version 4.1 (Waters, Manchester, UK).22, 23 The resulting data from the UPLC-QTOFMS platforms were subject to multivariate statistical analyses to establish characteristic metabolomic profiles associated with obesity before and after dietary intervention. For details, see Materials and Methods in the Supporting Information.

For GC-TOFMS generated data, the acquired MS files were analyzed according to our previous published work.21, 22 Briefly, the data generated in the GC-TOFMS instrument were analyzed by the ChromaTOF (v4.33, Leco Co., CA, USA). The resulting three dimension data set including sample information, peak retention time and peak intensities were subject to multivariate statistical analyses to establish characteristic metabolomic profiles associated with obesity before and after dietary intervention. For details, see Methods in the Supporting Information.

For UPLC-QTOFMS generated data, compound annotation was performed by comparing the accurate mass (m/z) and retention time (Rt) of reference standards and the accurate mass of compounds obtained from the web-based resources such as the Human Metabolome Database (www.hmdb.ca). For GC-TOFMS generated data, compound annotation was processed by comparing the mass fragments and Rt with the reference standards or mass fragments with NIST 05 Standard mass spectral databases in NIST MS search 2.0 (NIST, Gaithersburg, MD) software using a similarity of more than 70%.

The annotated metabolites in the two data sets resulting from UPLC-QTOFMS and GC-TOFMS were combined into a new data set for further statistical analysis by uni- and multivariate statistical methods. The combined data set was imported into SIMCA-P+ 12.0 software package (Umetrics, Umeå, Sweden). Principle component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA) were carried out to visualize the metabolic alterations between obese subjects and healthy controls. The metabolic effects of VLCD intervention were also investigated. In this study, the default 7-round cross-validation was applied with 1/7th of the samples being excluded from the mathematical model in each round in order to guard against over-fitting. The variable importance in the projection (VIP) values of all of the peaks from the 7-fold cross-validated OPLS-DA model were taken as a coefficient for peak selection. VIP ranks the overall contribution of each variable to the OPLS-DA model, and those variables with VIP > 1.0 are considered relevant for group discrimination 24 In addition to the multivariate statistical method, the Mann Whitney U test was also applied to measure the significance of each metabolite. Metabolites with both multivariate and univariate statistical significance (VIP > 1 and p < 0.05) were considered potential markers responsible for the differentiation of obesity subjects from healthy controls or obesity subjects before and after VLCD intervention.

Pearson correlation analysis was made to evaluate the relation of metabolite change along different time points (baseline, 4 and 8 weeks after VLCD intervention) versus the change of BMI, serum glucose level and insulin sensitivity, giving a value ranging from 1.0 (maximum positive correlation) to −1 (maximum anticorrelation) and 0 (no correlation). More specifically, using the ratio of metabolite change over baseline (FC value), we also evaluated the correlation of this ratio versus the corresponding change of BMI, serum glucose level and insulin sensitivity at 4 weeks and 8 weeks within the sub groups of female and male participants, correspondingly.

Results

Demographics and Clinical Characteristics.

Demographics and clinical characteristics of healthy control and obese human are shown in Table 1. Significant difference was observed in age (P < 0.05) between the obese and healthy subject groups (mean age: 28 y vs. 32 y, respectively). As expected, obese subjects were heavier and had higher WHR than healthy controls. The mean BMI of obese subjects was 32.58 kg/m2, whereas the mean BMI was 21.29 kg/m2 for healthy controls.

Obese subjects had higher levels of insulin (p < 0.05), HOMA-IR (p < 0.05), and blood glucose (p < 0.05) than healthy controls (Table 1).

As shown in Table 1, after the VLCD intervention, the BMI was significantly reduced from 32.58 kg/m2 to 30.59 kg/m2 (p < 0.05) at week 4 and further reduced to 29.88 kg/m2 (p < 0.01) at week 8. Similarly, significant reductions in fasting insulin (FINS) and 2h postprandial insulin (2h INS) were observed (all P values < 0.05). However, the change of fasting plasma glucose (FPG, P = 0.503) and 2h postprandial plasma glucose (2h PG, P = 0.079) was not statistic significant at the end of the study. HOMA-IR and 2h HOMA-IR significantly improved with a 2-fold reduction (all P values < 0.05), respectively.

Serum Metabolite Profiles of Obese and Healthy Control Subjects.

A total of 113 metabolites were annotated from the detected spectral features from GC-TOFMS and UPLC-QTOFMS using reference standards as well as the available database (NIST library 2005 and HMDB). Peak intensity comparison of the differentially expressed metabolite levels between control and obese subjects before and after VLCD intervention were summarized in Table 2.

Table 2.

List of Serum Differential Metabolites in Obese Subjects Relative to Healthy Controls, Obese Subjects after VLCD Intervention Relative to Controls, and Obese Subjects after VLCD Intervention Relative to Obese Subjects.

Compound Class Obesity-control Obesity_VLCD 4 wk-control Obesity_VLCD 8 wk-control Obesity_VLCD 4wk-Obesity Obesity_VLCD 8 wk-Obesity
VIP FC p VIP FC p VIP FC p VIP FC p VIP FC p
Carnitine Alkylamines 1.14 1.25 2.26E-04 0.57 1.12 3.40E-02 0.90 1.16 4.45E-03
Ethylamine Alkylamines 2.10 0.73 6.42E-12 2.33 0.74 1.87E-11 2.32 0.74 2.36E-11
Ethanolamine Alkylamines 1.94 0.62 3.25E-10 2.17 0.62 2.03E-10 2.22 0.61 1.52E-10
Butylamine Alkylamines 1.38 0.88 1.97E-08 2.09 0.89 9.32E-09 1.99 0.89 6.72E-08
Spermidine Alkylamines 1.78 0.84 2.88E-08 1.97 0.85 6.79E-08 1.89 0.85 6.13E-07
Glutamate Amino acid 1.74 2.19 6.56E-10 1.49 1.85 4.41E-07 1.58 1.64 1.47E-05 1.08 0.84 2.79E-02 1.82 0.75 3.40E-03
Alanine Amino acid 1.18 1.34 2.73E-05 0.24 1.06 5.67E-01 0.28 1.06 3.64E-01 1.79 0.79 1.89E-04 1.76 0.79 1.32E-03
Homocysteine Amino acid 1.10 0.77 1.41E-03 1.59 0.70 3.78E-05 1.48 0.70 1.13E-04
Homophenylalanine Amino acid 1.07 1.51 4.42E-05 1.22 2.10 3.14E-06 0.95 1.73 1.55E-02
Phenylalanine Amino acid 1.88 1.59 5.70E-10 1.78 1.49 6.02E-08 1.72 1.44 1.85E-06
Aspartate Amino acid 1.81 1.60 5.32E-10 1.84 1.60 1.28E-08 1.80 1.56 2.52E-07
Cystine Amino acid 1.81 2.58 4.94E-11 1.59 1.90 1.07E-06 1.95 2.57 1.58E-08 1.45 0.74 2.82E-03
Valine Amino acid 1.39 1.32 6.36E-06 1.35 1.34 7.56E-06 1.16 1.27 1.23E-03
Leucine Amino acid 1.16 1.25 1.22E-04 1.15 1.26 5.17E-04 1.00 1.21 2.38E-03
Isoleucine Amino acid 1.05 1.27 4.27E-04 1.02 1.28 2.05E-03 1.00 1.24 4.45E-03
N-acetylneuraminate Carbohydrates 1.43 1.51 7.43E-06 1.11 1.38 4.76E-03 1.25 1.29 2.95E-04
Ribose Carbohydrates 1.22 0.51 2.45E-06 0.43 0.83 8.75E-02 0.31 1.15 9.81E-01 1.42 1.64 1.77E-03 2.21 2.26 2.21E-05
Threonate Carbohydrates 1.83 2.73 6.55E-09 1.83 2.81 1.89E-10 1.95 2.98 2.79E-09
Ribitol Carbohydrates 1.37 0.40 5.69E-04 1.70 0.36 6.90E-05 1.56 0.39 4.90E-04
Mannose Carbohydrates 1.28 1.34 4.21E-05 2.03 1.68 9.29E-10 1.91 1.60 3.17E-08 1.78 1.25 2.22E-04 1.59 1.19 1.15E-02
Pseudo uridine Carbohydrates 1.02 1.39 1.90E-04 0.86 1.28 1.29E-02 1.22 1.31 1.46E-03
Succinate Carboxylic Acids 1.77 3.15 1.99E-12 1.99 2.48 2.70E-11 1.74 2.17 7.21E-09 1.17 0.79 3.22E-02 1.73 0.69 4.00E-04
5-oxoproline Carboxylic Acids 1.75 1.47 7.97E-09 1.32 1.40 2.85E-05 1.42 1.33 6.82E-05
Malic acid Carboxylic Acids 1.39 1.41 4.41E-06 1.25 1.41 2.79E-04 0.66 1.31 1.65E-02
Myo-Inositol Cyclic Alcohols 1.35 1.29 1.51E-05 1.05 1.28 8.32E-03 1.02 1.22 7.47E-03
cis-5,8,11,14,17-Eicosapentaenoate Fatty acid 1.65 2.13 1.67E-06 1.02 1.98 8.06E-03 1.33 1.83 9.71E-05 1.38 0.93 4.98E-02
cis-5,8,11,14-eicosatetraenoate Fatty acid 1.33 1.75 1.34E-07 0.99 1.38 1.04E-02 1.25 1.37 9.71E-05
Palmitoleate Fatty acid 0.99 1.51 5.69E-04 0.44 1.19 3.41E-01 0.80 1.26 1.45E-02 1.12 0.79 2.56E-02 1.13 0.84 1.95E-01
Elaidate Fatty acid 1.19 1.48 9.29E-05 0.62 1.24 9.37E-02 0.89 1.24 8.32E-03 1.42 0.84 6.27E-02
Palmitate Fatty acid 1.19 1.43 5.86E-05 0.58 1.22 1.50E-01 1.15 1.29 1.59E-04
Stearate Fatty acid 1.16 1.27 2.07E-04 0.54 1.11 1.50E-01 0.19 1.03 1.00E+00 1.31 0.87 1.05E-02 1.84 0.81 3.86E-04
Pentadecanoate Fatty acid 1.28 1.76 2.72E-05 0.84 1.43 3.59E-02 1.43 1.76 1.72E-05 1.06 0.81 2.51E-02
Ethylhexanoate Fatty acid 1.25 0.73 2.60E-05 1.11 0.78 2.05E-03 1.46 0.68 4.07E-05
Fumarate Fatty Acids 1.02 0.88 2.38E-03 1.64 0.82 3.88E-06 1.37 0.84 5.36E-04 1.11 0.93 6.60E-02
Oleamide Fatty Amides 1.63 0.13 2.88E-06 0.98 0.40 1.53E-04 0.17 1.12 9.90E-01 2.53 8.92 4.43E-04
2-hydroxybutyrate Hydroxy Acids 1.60 1.70 5.78E-07 1.84 2.86 2.34E-10 1.42 2.00 3.13E-05 2.24 1.68 1.27E-05
2,3-Dihydroxybutanoate Hydroxy Acids 1.23 1.65 4.85E-05 1.77 2.49 1.99E-08 1.37 1.88 1.02E-04 1.93 1.51 6.22E-04
Adenine Imidazopyrimidines 1.17 0.42 5.07E-05 0.02 0.97 3.30E-01 0.45 0.73 1.65E-02 1.61 2.30 7.32E-04
Xanthine Imidazopyrimidines 1.51 1.56 2.20E-06 1.06 1.44 1.32E-03 0.74 1.24 3.03E-02 1.68 0.80 5.39E-03
Uric acid Imidazopyrimidines s 1.04 1.57 4.83E-04 1.02 1.50 1.91E-03 1.01 1.50 6.00E-03
4-hydroxy-3-methoxymandelate Phenols 1.22 0.87 3.27E-02 0.81 0.94 3.52E-01 1.09 1.00 6.28E-01 1.29 1.16 1.34E-02
Nicotinamide Pyridines 1.57 1.33 5.78E-07 0.90 1.15 1.33E-02 1.18 1.20 2.58E-03 1.69 0.86 7.67E-04 1.14 0.91 1.86E-02
chenodeoxycholate Steroids 1.11 0.45 1.52E-03 0.51 0.73 1.17E-01 0.31 0.85 8.09E-01 2.12 1.91 1.90E-04
Taurine Sulfonic Acids 1.22 1.32 1.06E-04 1.18 1.27 1.32E-03 1.48 1.37 2.81E-05
2-Aminobutyrate Amino acid 2.74 1.64 2.79E-07
Proline Amino acid 2.37 0.73 1.86E-06 1.71 0.81 7.12E-04
Ornithine Amino acid 1.82 0.68 1.60E-04 1.70 0.71 1.81E-03
Tryptophan Amino acid 1.48 0.84 5.01E-03
Methionine Amino acid 1.18 0.81 5.01E-03
Threonine Amino acid 1.36 0.88 1.10E-02
Theanine Amino acid 1.31 1.36 1.75E-02 0.71 1.18 1.98E-01
cis-11,14-eicosadienoate Fatty acid 2.05 0.83 3.64E-02
cis-11,14,17-Eicosatrienoate Fatty acid 1.04 0.75 2.75E-05
Arachidonate Fatty acid 1.35 1.17 1.51E-02
Acetyl-carnitine Fatty acid esters 2.21 1.47 6.89E-06
3-Hydroxybutyrate Hydroxy Acids 2.32 6.33 3.78E-07
p-cresol Phenols 2.18 2.54 2.95E-05 1.96 1.96 2.20E-03
3-Aminophenol Phenols 1.23 0.62 3.16E-02
Threitol Sugar Alcohols 1.80 0.76 2.01E-04 1.26 0.83 1.18E-02
Urea Ureas 1.13 1.19 1.55E-02 1.44 1.23 3.59E-02
Glyceraldehyde Alcohols 2.30 1.35 7.86E-04
Glycerol Sugar Alcohols 1.36 0.81 6.92E-03
Open in a new tab
a

The metabolites responsible for the differentiation of metabolic profiles of obese subjects; or obese subjects with 4 weeks intervention; or obese subjects with 8 weeks intervention, from controls and metabolic profiles of obese subjects with 4 weeks intervention; or obese subjects with 8 weeks intervention from obese subjects were obtained using a univariate statistical analysis, Mann Whitney U test. The corresponding fold change shows how these selected differential metabolites varied in the obese subjects before and after VLCD intervention from those of normal controls or in the obese subjects after VLCD intervention from those of obese subjects. VIP was obtained from OPLS-DA with a threshold of 1.0.

To distinguish healthy controls from obese subjects at baseline as well as after VLCD intervention, principal component analysis (PCA) and orthogonal partial least squares-discriminant analysis (OPLS-DA) analysis were performed with the 113 metabolites generated (Figure 1). There appears to be a separation between healthy controls and obese subjects, reflecting the pathophysiological variations of obesity. Notably, the metabolic patterns of obese subjects deviate from baseline after 8 weeks VLCD intervention and the metabolic pattern is still distinct from the healthy controls. Analogously, distinct separation was seen among the metabolite profiles of the obese subjects before and after VLCD intervention, indicative of the effects of dietary intervention on health pathology (Figure 2 and supplementary Figure S1).

Figure 1.

Figure 1.

Open in a new tab

(A) PCA and (B) OPLS-DA scores plot constructed with the 113 annotated metabolites.

Figure 2.

Figure 2.

Open in a new tab

OPLS-DA scores plot constructed with the 113 annotated metabolites.

Based on VIP value (> 1) and Mann−Whitney U test p value (< 0.05), a total of 44 metabolites were significantly altered in obese subjects compared to healthy controls (Table 2) and the results were summarized as Venn diagram exhibiting the commonly and specifically affected metabolites (Figure 3B). These metabolites represent key metabolic pathways involving amino acid metabolism, fatty acid metabolism, lipid metabolism, carbohydrates metabolism and TCA cycle (Figure 3C).

Figure 3.

Figure 3.

Open in a new tab

(A) Heatmap shows changes in metabolites compared to healthy control at obesity subjects, obesity subjects after VLCD intervention at week four and eight. Shades of yellow and blue represent fold increase and fold decrease of a metabolite, respectively, in obesity subjects, obesity subjects after 4 weeks VLCD intervention or obesity subjects after 8 weeks VLCD intervention relative to healthy controls (see color scale); (B) metabolic pathways being affected by VLCD intervention; and (C) Venn diagramm exhibiting the commonly and specifically affected metabolites.

Beneficial Effects of VLCD Intervention in Obese Subjects

The variations of differential metabolites listed in Table 2 upon obesity were investigated in obesity subjects after VLCD intervention, and as a result, not all the metabolites were attenuated or normalized. Metabolomic response profiles of representative metabolites listed in Table 2 upon obesity and obesity after VLCD intervention was depicted as a heatmap in Figure 3A. The heatmap generated indicates less significant fluctuation of metabolite levels (in fold change, relative to healthy control) in obesity after VLCD intervention, suggesting that VLCD could attenuate the metabolic perturbation in obese subjects.

The concentration of carnitine, fatty acids including cis-5,8,11,14,17-eicosapentaenoate, cis-5,8,11,14-eicosatetraenoate, palmitoleate, elaidate, palmitate, and stearate, succinate, 5-oxoproline, xanthine, urate, nicotinamide, chenodeoxycholate, oleamide, 2-hydroxybutyrate, N-acetylneuraminate, ribose, glutamate, and alanine were altered in obese subjects. After VLCD intervention, however, the levels of above metabolites were altered in less significant level, suggesting that VLCD can attenuate the metabolic alteration in obesity.

We further selected the differentially expressed serum metabolites before and after 4 weeks or 8 weeks VLCD intervention in the obese subjects based on the VIP values (VIP > 1) by one predictive component and two orthogonal component OPLS-DA models (R2X = 0.293, R2Y = 0.772, Q2(cum) = 0.591; R2X = 0.293, R2Y = 0.783, Q2(cum) = 0.561), respectively (Supplementary Figure S1). A list of differential metabolites (Table 2) including 2-aminobutyrate, proline, ornithine, tryptophan, methionine, threonine, theanine, cis-11,14-eicosadienoate, cis-11,14,17-eicosatrienoate, arachidonate, acetyl-carnitine, 3-hydroxybutyrate, p-cresol, 3-aminophenol, threitol, urea, glutamate, alanine, cysteine, ribose, mannose, succinate, cis-5,8,11,14,17-eicosapentaenoate, palmitoleate, stearate, pentadecanoate, fumarate, 2-hydroxybutyrate, 2,3-dihydroxybutanoate, adenine, and nicotinamide was identified after 4 weeks VLCD intervention and proline, ornithine, tryptophan, threonine, p-cresol, threitol, urea, glutamate, alanine, ribose, mannose, succinate, palmitoleate, elaidate, stearate, oleamide, xanthine, 4-hydroxy-3-methoxymandelate, nicotinamide, chenodeoxycholate, glyceraldehyde, and glycerol after 8 weeks VLCD intervention. Most of them are different from those differential metabolites in obese subjects relative to healthy controls.

Correction of Metabolite Change along Different Time Points (Baseline, 4 Weeks and 8 Weeks after VLCD Intervention) versus the Change of BMI, Plasma Glucose Level and Insulin Sensitivity

A correlation analysis was performed among the 44 differential metabolites (Figure 4 and S2), which revealed a wide range of correlation coefficients among the interactions between time (baseline, 4 and 8 weeks after VLCD intervention), treatment (VLCD intervention) and clinical parameters (the change of BMI, serum glucose level and insulin sensitivity), ranging from 1.0 (maximum positive correlation) to −0.5 (maximum anticorrelation) and 0 (no correlation, see color bar scale in Figure 4). Figure 4 and Supplementary Figure S2 illustrate that several high positive (dark red and red regions) or negative (blue regions) correlations were observed among several metabolites among male and female participants. From the correlation difference matrix, succinate, alanine, and fumarate were positively correlated with changes of BMI, FPG and FINS in females, but with no correlation in males. Similarly, glycochenodeoxycholic acid, methionine, uric acid, leucine, and isoleucine were positively correlated with changes of BMI in male participants, but with no correlation in females. Glalactose, arabinofuranose, and thymine were positively correlated with changes of FPG and FINS in both males and females.

Figure 4.

Figure 4.

Open in a new tab

Heatmap shows interactions between time (metabolite change from baseline, 4 weeks, and then 8 weeks), VLCD intervention (4 weeks and 8 weeks) and clinical parameters. Shades of red and blue represent positive correlation and negative correlation, respectively (see color scale).

Using the ratio of metabolite change over baseline (FC value), we also evaluated the correlation of this ratio versus the corresponding change of BMI, serum glucose level and insulin sensitivity at 4 weeks and 8 weeks within the sub groups of female and male participants, correspondingly. As shown in Figure 5, fatty acids including cis-5,8,11,14,17-eicosapentaenoate, cis-5,8,11,14-eicosatetraenoate, palmitoleate, elaidate, palmitate, myristate, and linoleate were positively correlated with the changes of BMI in female after 8 weeks VLCD intervention, but with no correlation in males. The change of BMI was positively correlated with the changes of FINS in males after 8 weeks VLCD intervention but with less correlation in females. The change of cis-5,8,11,14,17-eicosapentaenoic acid was negatively correlated with the change of PFG and FINS in females after 4 weeks VLCD intervention, but this correlation became positive after 8 weeks VLCD intervention. The changes of oleamide, 2-hydroxybutyrate and 2,3-dihydroxybutanoate were negatively correlated with the change of BMI, FPG and FINS in both males and females after 4 weeks VLCD intervention, which were become more negative after 8 weeks VLCD intervention. BCAAs, leucine, isoleucine and valine were positively correlated with the changes of BMI, FPG and FINS both in males and females after 4 or 8 weeks VLCD intervention.

Figure 5.

Figure 5.

Open in a new tab

Heatmaps show the correlation of the change of metabolite with baseline (A) after 4 weeks VLCD intervention and (B) after 8 weeks VLCD intervention versus the changes in BMI, FPG, and FINS. Shades of red and blue represent positive correlation and negative correlation, respectively; in the change of metabolite in obesity subjects after 4 weeks VLCD intervention or obesity subjects after 8 weeks VLCD intervention relative to the changes of clinical parameters (see color scale).

Metabolite Markers Associated With Age

Because there is a statistically significant difference in age between obese subjects and healthy controls, we tried to identify metabolite markers associated with age among the subjects in the control group. We first established the PCA scores plot (three component, R2X = 0.400, Q2(cum) = 0.108, Supplementary Figure S3) of all the healthy controls and there’s no clear separation among the healthy controls with different ages. We then compared all the metabolites between the subjects of older ages (> 29 yrs. old) and the younger subjects (< 25 yrs. old) and one metabolite, tocopherol (vitamin E) of statistical significance (p < 0.05) was removed from the list of potential markers.

Discussion

In the present study, we applied a comprehensive metabonomics approach to understand the metabolic differences between obese and healthy control subjects and determine how those metabolic profiles were impacted by VLCD intervention. Several obesity-related changes described herein confirm prior studies, including the higher levels of blood glucose, insulin and HOMA-IR in obese subjects compared to healthy controls.12 The OPLS-DA analysis of healthy controls and obese subjects revealed that obese subjects had a clearly distinct metabolic profile from healthy controls. Several metabolites associated with amino acid metabolism, fatty acid metabolism, lipid metabolism, carbohydrate metabolism and TCA cycle were altered, and changes in these metabolic profiles contributed to the difference between healthy controls and obese subjects (Table 2). These results were consistent with previously published metabolomic studies analyzing healthy obese and control subjects, where several key metabolic pathways, including BCAA metabolism, fatty acid metabolism, bile acid metabolism, and gut microbial-host co-metabolism were significantly altered in association with the obese phenotype.13, 14 VLCD intervention resulted in a global metabolite alteration in obese subjects, leading to an attenuated metabolite perturbation in obese subjects, suggesting that VLCD could attenuate the obesity induced metabolic perturbation. Amino acids are well known to stimulate the endogenous release of glucoregulatory hormones and might thereby modulate glucose metabolism.25, 26 Interestingly, ingestion of VLCD actually decreases blood glucose level in obesity participants. Changes of amino acids were positively correlated with the changes in BMI, FPG and FINS. The decreased BMI, FPG and FINS was consistent with the results of the decreased levels of amino acids after VLCD intervention. However, not all the differentially expressed metabolites in obese subjects were attenuated or normalized by VLCD intervention. Therefore, a list of differential metabolites (Table 2) responsible for the separation between obese subjects before and after VLCD intervention was identified. These “VLCD-induced” markers are different from those differential metabolites in obese subjects relative to the healthy controls.

Consistent with the results presented by Newgard et al.,13 we found that key compounds of fatty acid synthesis and oxidation (carnitine) were significantly increased in obese subjects. Fatty acids are an important energy source in the body and provide energy through β-oxidation. Research has shown that obesity and diabetes are related to the decreased fatty acid oxidation capacity, and as a result, excessive fatty acids will accumulate in the body as triglycerides, thereby causing lipotoxicity. Many reports have revealed that fatty liver, adipose tissue, and pancreatic fat accumulation are directly related to insulin resistance.27–29 Serum fatty acid composition may modulate insulin action, and increased serum fatty acid concentrations are known to impair glucose metabolism, potentially causing diabetes.30 Consistent with the previous report by Kim et al.,14 cis-5,8,11,14,17-eicosapentaenoic acid, cis-5,8,11,14-eicosatetraenoic acid, palmitoleic acid, elaidic acid, stearic acid, palmitic acid, pentadecanoic acid, were found to be significantly increased in obese subjects compared to healthy controls, representing fatty acids that circulate as triglycerides or other esterified species. These findings are consistent with the strong increase in triglyceride levels in obese subjects. It was reported that hypocaloric diets with VLC (12% energy) in obese subjects lead to greater increases in markers of lipolysis (ketones and free fatty acids) together with greater decreases of markers of lipogenesis and deposition of body fat (leptin) when compared to low-fat (56% energy as CHO) hypocaloric diets.31 As a product of fat lipolysis and oxidation, 3-hydroxybutyrate, was of 6.33-fold increase at week 4 and subsequently decreased in week 8, with its concentration remaining significantly higher than that at baseline. The changes of 3-hydroxybutyrate was negatively correlated with the change of BMI, FPG and FINS in both males and females after 4 or 8 weeks VLCD intervention (Supplementary Figure S2), which was consistent with the results of the increased 3-hydroxybutyrate level and the decreased clinical parameters after VLCD intervention. A study reported that fatty acids significantly increased in the first 7 days but decreased in week 8 in diabetic patients who participated in an 8-week VLCD intervention.32 Considering the changes of fatty acids and ketone body together, the changes might reflect the process of fat burning during the 8-week period, leading to the significant weight loss.

The changes in amino acid metabolism have also been reported for obese people.33, 34 Of particular interest, the physiological effects of BCAA have been noted in various aspects of energy metabolism, including the stimulation of insulin secretion in pancreatic beta cells 35 and the control of appetite through the mTOR pathway in hypothalamic neurons.36 It is therefore likely that in the future, BCAA will be used clinically as a marker of insulin resistance. Emerging evidence suggests BCAA could be a marker for obesity and combined with phenylalanine and tyrosine could be used for the prediction of diabetes risks.13, 37 The higher BCAA levels on obese subjects may reflect an overload of BCAA catabolism in obese subjects.13 However, the levels of BCAA in obese subjects did not change significantly after VLCD intake, which indicates that the VLCD is unlikely affecting dietary intake or plasma concentrations of BCAAs. BCAAs, leucine, isoleucine and valine were positively correlated with the changes of BMI, FPG and FINS both in males and females after 4 or 8 weeks VLCD intervention, which was consistent with the decreased BCAAs level after VLCD intervention, although no significant changes were found. The gut microbiota has recently been suggested to contribute to the development of obesity. The composition of the gut microbiota has been shown to differ in lean and obese humans and animals and to change rapidly in response to dietary factors.38 Interestingly, there is direct evidence that changing the amount and/or type of carbohydrate over periods of up to four weeks has a profound and rapid influence on the composition of the gut microbiota and its metabolic products in adult human volunteers.39–42 Some metabolites including bile acids, phenolic, benzoyl, and phenyl derivatives and lipids that regulate host-microbiota interactions are believed to be involved in gut microbiota-host co-metabolism.43 Four or eight weeks VLCD intervention clearly altered the metabolites involved in gut microbiota which supports the reported findings of altered gut microbiota after diet intervention.

The aim of this study was to determine metabolic differences between obese and healthy human subjects and investigate the effect of VLCD intervention on serum metabolic profiles in obese subjects by metabonomic approach. However, there are several limitations in the current MS-based study. First, the relatively small size used in this study may not be sufficiently large to detect all diet-associated metabolic changes. Second, being a short-term designed study, the evidence of the long-term effects of VLCD was unclear. Longer-term and further studies are needed to investigate the effects and to reveal the underlying mechanisms of VLCD intervention on weight loss and weight maintenance, as well as on glucose prevention in the future.

Conclusion

In conclusion, we identified comprehensive metabolic shifts in obese subjects including altered fatty acid metabolism, amino acid metabolism, carbohydrate metabolism and TCA cycle. VLCD is able to attenuate metabolic perturbation in obese subjects, characterized by the less significant metabolic alterations. It also appears that VLCD induced significant metabolic alterations independent of the obesity induced metabolic changes. These VLCD-induced alterations include a panel of metabolites that are involved in inflammation and oxidation processes, suggesting that VLCD exerts the beneficial effects associating with the anti-inflammatory and antioxidant mechanisms in addition to the attenuation of obesity-induced metabolic perturbation.

Supplementary Material

SI

Acknowledgment

This study was financially supported by Drug Innovation Program of National Science and Technology (Project No. 2011ZX09307-001-02).

Footnotes

Supporting Information

Supplementary methods, tables, and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

References

  • 1.James PT; Rigby N; Leach R, The obesity epidemic, metabolic syndrome and future prevention strategies. Eur J Cardiovasc Prev Rehabil 2004, 11 (1), 3–8. [DOI] [PubMed] [Google Scholar]
  • 2.Van Gaal LF; Mertens IL; De Block CE, Mechanisms linking obesity with cardiovascular disease. Nature 2006, 444 (7121), 875–80. [DOI] [PubMed] [Google Scholar]
  • 3.Frigolet ME; Ramos Barragan VE; Tamez Gonzalez M, Low-carbohydrate diets: a matter of love or hate. Ann Nutr Metab 2011, 58 (4), 320–34. [DOI] [PubMed] [Google Scholar]
  • 4.McAuley KA; Hopkins CM; Smith KJ; McLay RT; Williams SM; Taylor RW; Mann JI, Comparison of high-fat and high-protein diets with a high-carbohydrate diet in insulin-resistant obese women. Diabetologia 2005, 48 (1), 8–16. [DOI] [PubMed] [Google Scholar]
  • 5.Foster GD; Wyatt HR; Hill JO; McGuckin BG; Brill C; Mohammed BS; Szapary PO; Rader DJ; Edman JS; Klein S, A randomized trial of a low-carbohydrate diet for obesity. The New England journal of medicine 2003, 348 (21), 2082–90. [DOI] [PubMed] [Google Scholar]
  • 6.Boden G; Sargrad K; Homko C; Mozzoli M; Stein TP, Effect of a low-carbohydrate diet on appetite, blood glucose levels, and insulin resistance in obese patients with type 2 diabetes. Annals of internal medicine 2005, 142 (6), 403–11. [DOI] [PubMed] [Google Scholar]
  • 7.Weigle DS; Breen PA; Matthys CC; Callahan HS; Meeuws KE; Burden VR; Purnell JQ, A high-protein diet induces sustained reductions in appetite, ad libitum caloric intake, and body weight despite compensatory changes in diurnal plasma leptin and ghrelin concentrations. Am J Clin Nutr 2005, 82 (1), 41–8. [DOI] [PubMed] [Google Scholar]
  • 8.Sargrad KR; Homko C; Mozzoli M; Boden G, Effect of high protein vs high carbohydrate intake on insulin sensitivity, body weight, hemoglobin A1c, and blood pressure in patients with type 2 diabetes mellitus. Journal of the American Dietetic Association 2005, 105 (4), 573–80. [DOI] [PubMed] [Google Scholar]
  • 9.Luscombe ND; Clifton PM; Noakes M; Parker B; Wittert G, Effects of energy-restricted diets containing increased protein on weight loss, resting energy expenditure, and the thermic effect of feeding in type 2 diabetes. Diabetes Care 2002, 25 (4), 652–7. [DOI] [PubMed] [Google Scholar]
  • 10.Golay A; Allaz AF; Morel Y; deTonnac N; Tankova S; Reaven G, Similar weight loss with low- or high-carbohydrate diets. American Journal of Clinical Nutrition 1996, 63 (2), 174–178. [DOI] [PubMed] [Google Scholar]
  • 11.Gannon MC; Nuttall FQ; Saeed A; Jordan K; Hoover H, An increase in dietary protein improves the blood glucose response in persons with type 2 diabetes. American Journal of Clinical Nutrition 2003, 78 (4), 734–741. [DOI] [PubMed] [Google Scholar]
  • 12.Gu YJ; Yu HY; Li YH; Ma XJ; Lu JX; Yu WH; Xiao YF; Bao YQ; Jia WP, Beneficial effects of an 8-week, very low carbohydrate diet intervention on obese subjects. Evidence-Based Complementary and Alternative Medicine 2013, 2013, 760804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Newgard CB; An J; Bain JR; Muehlbauer MJ; Stevens RD; Lien LF; Haqq AM; Shah SH; Arlotto M; Slentz CA; Rochon J; Gallup D; Ilkayeva O; Wenner BR; Yancy WS Jr.; Eisenson H; Musante G; Surwit RS; Millington DS; Butler MD; Svetkey LP, A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab 2009, 9 (4), 311–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Kim JY; Park JY; Kim OY; Ham BM; Kim HJ; Kwon DY; Jang Y; Lee JH, Metabolic profiling of plasma in overweight/obese and lean men using ultra performance liquid chromatography and Q-TOF mass spectrometry (UPLC-Q-TOF MS). J Proteome Res 2010, 9 (9), 4368–75. [DOI] [PubMed] [Google Scholar]
  • 15.Pietilainen KH; Naukkarinen J; Rissanen A; Saharinen J; Ellonen P; Keranen H; Suomalainen A; Gotz A; Suortti T; Yki-Jarvinen H; Oresic M; Kaprio J; Peltonen L, Global transcript profiles of fat in monozygotic twins discordant for BMI: pathways behind acquired obesity. PLoS Med 2008, 5 (3), e51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Adams SH, Emerging perspectives on essential amino acid metabolism in obesity and the insulin-resistant state. Adv Nutr 2011, 2 (6), 445–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Gibney MJ; Walsh M; Brennan L; Roche HM; German B; van Ommen B, Metabolomics in human nutrition: opportunities and challenges. Am J Clin Nutr 2005, 82 (3), 497–503. [DOI] [PubMed] [Google Scholar]
  • 18.Lankinen M; Schwab U; Gopalacharyulu PV; Seppanen-Laakso T; Yetukuri L; Sysi-Aho M; Kallio P; Suortti T; Laaksonen DE; Gylling H; Poutanen K; Kolehmainen M; Oresic M, Dietary carbohydrate modification alters serum metabolic profiles in individuals with the metabolic syndrome. Nutr Metab Cardiovasc Dis 2010, 20 (4), 249–57. [DOI] [PubMed] [Google Scholar]
  • 19.Moazzami AA; Bondia-Pons I; Hanhineva K; Juntunen K; Antl N; Poutanen K; Mykkanen H, Metabolomics reveals the metabolic shifts following an intervention with rye bread in postmenopausal women--a randomized control trial. Nutr J 2012, 11, 88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Laferrere B; Reilly D; Arias S; Swerdlow N; Gorroochurn P; Bawa B; Bose M; Teixeira J; Stevens RD; Wenner BR; Bain JR; Muehlbauer MJ; Haqq A; Lien L; Shah SH; Svetkey LP; Newgard CB, Differential metabolic impact of gastric bypass surgery versus dietary intervention in obese diabetic subjects despite identical weight loss. Sci Transl Med 2011, 3 (80), 80re2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Chen T; Xie G; Wang X; Fan J; Qiu Y; Zheng X; Qi X; Cao Y; Su M; Xu LX; Yen Y; Liu P; Jia W, Serum and urine metabolite profiling reveals potential biomarkers of human hepatocellular carcinoma. Mol Cell Proteomics 2011, 10 (7), M110 004945. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Qiu Y; Cai G; Su M; Chen T; Zheng X; Xu Y; Ni Y; Zhao A; Xu LX; Cai S; Jia W, Serum metabolite profiling of human colorectal cancer using GC-TOFMS and UPLC-QTOFMS. J Proteome Res 2009, 8 (10), 4844–50. [DOI] [PubMed] [Google Scholar]
  • 23.Xie G; Ye M; Wang Y; Ni Y; Su M; Huang H; Qiu M; Zhao A; Zheng X; Chen T; Jia W, Characterization of pu-erh tea using chemical and metabolic profiling approaches. J Agric Food Chem 2009, 57 (8), 3046–54. [DOI] [PubMed] [Google Scholar]
  • 24.Jansson J; Willing B; Lucio M; Fekete A; Dicksved J; Halfvarson J; Tysk C; Schmitt-Kopplin P, Metabolomics reveals metabolic biomarkers of Crohn’s disease. PLoS One 2009, 4 (7), e6386. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Floyd JC Jr.; Fajans SS; Conn JW; Knopf RF; Rull J, Stimulation of insulin secretion by amino acids. The Journal of clinical investigation 1966, 45 (9), 1487–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Ohneda A; Parada E; Eisentraut AM; Unger RH, Characterization of response of circulating glucagon to intraduodenal and intravenous administration of amino acids. The Journal of clinical investigation 1968, 47 (10), 2305–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Cheung O; Sanyal AJ, Recent advances in nonalcoholic fatty liver disease. Curr Opin Gastroenterol 2010, 26 (3), 202–8. [DOI] [PubMed] [Google Scholar]
  • 28.Muoio D, Revisiting the connection between intramyocellular lipids and insulin resistance: a long and winding road. Diabetologia, 1–4. [DOI] [PubMed] [Google Scholar]
  • 29.Tushuizen ME; Bunck MC; Pouwels PJ; Bontemps S; van Waesberghe JH; Schindhelm RK; Mari A; Heine RJ; Diamant M, Pancreatic fat content and beta-cell function in men with and without type 2 diabetes. Diabetes Care 2007, 30 (11), 2916–21. [DOI] [PubMed] [Google Scholar]
  • 30.Wang L; Folsom AR; Zheng ZJ; Pankow JS; Eckfeldt JH, Plasma fatty acid composition and incidence of diabetes in middle-aged adults: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Clin Nutr 2003, 78 (1), 91–8. [DOI] [PubMed] [Google Scholar]
  • 31.Volek JS; Phinney SD; Forsythe CE; Quann EE; Wood RJ; Puglisi MJ; Kraemer WJ; Bibus DM; Fernandez ML; Feinman RD, Carbohydrate Restriction has a More Favorable Impact on the Metabolic Syndrome than a Low Fat Diet. Lipids 2009, 44 (4), 297–309. [DOI] [PubMed] [Google Scholar]
  • 32.Lim EL; Hollingsworth KG; Aribisala BS; Chen MJ; Mathers JC; Taylor R, Reversal of type 2 diabetes: normalisation of beta cell function in association with decreased pancreas and liver triacylglycerol. Diabetologia 2011, 54 (10), 2506–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Wijekoon EP; Skinner C; Brosnan ME; Brosnan JT, Amino acid metabolism in the Zucker diabetic fatty rat: effects of insulin resistance and of type 2 diabetes. Can J Physiol Pharmacol 2004, 82 (7), 506–14. [DOI] [PubMed] [Google Scholar]
  • 34.Kochhar S; Jacobs DM; Ramadan Z; Berruex F; Fuerholz A; Fay LB, Probing gender-specific metabolism differences in humans by nuclear magnetic resonance-based metabonomics. Anal Biochem 2006, 352 (2), 274–81. [DOI] [PubMed] [Google Scholar]
  • 35.Kalogeropoulo D; LaFave L; Schweim K; Gannon MC; Nuttall FQ, Leucine, when ingested with glucose, synergistically stimulates insulin secretion and lowers blood glucose. Metabolism-Clinical and Experimental 2008, 57 (12), 1747–1752. [DOI] [PubMed] [Google Scholar]
  • 36.Kahn BB; Myers MG, mTOR tells the brain that the body is hungry. Nature Medicine 2006, 12 (6), 615–617. [DOI] [PubMed] [Google Scholar]
  • 37.Wang TJ; Larson MG; Vasan RS; Cheng S; Rhee EP; McCabe E; Lewis GD; Fox CS; Jacques PF; Fernandez C; O’Donnell CJ; Carr SA; Mootha VK; Florez JC; Souza A; Melander O; Clish CB; Gerszten RE, Metabolite profiles and the risk of developing diabetes. Nat Med 2011, 17 (4), 448–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Tremaroli V; Backhed F, Functional interactions between the gut microbiota and host metabolism. Nature 2012, 489 (7415), 242–249. [DOI] [PubMed] [Google Scholar]
  • 39.Walker AW; Ince J; Duncan SH; Webster LM; Holtrop G; Ze XL; Brown D; Stares MD; Scott P; Bergerat A; Louis P; McIntosh F; Johnstone AM; Lobley GE; Parkhill J; Flint HJ, Dominant and diet-responsive groups of bacteria within the human colonic microbiota. Isme J 2011, 5 (2), 220–230. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Duncan SH; Belenguer A; Holtrop G; Johnstone AM; Flint HJ; Lobley GE, Reduced dietary intake of carbohydrates by obese subjects results in decreased concentrations of butyrate and butyrate-producing bacteria in feces. Appl Environ Microb 2007, 73 (4), 1073–1078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Russell WR; Gratz SW; Duncan SH; Holtrop G; Ince J; Scobbie L; Duncan G; Johnstone AM; Lobley GE; Wallace RJ; Duthie GG; Flint HJ, High-protein, reduced-carbohydrate weight-loss diets promote metabolite profiles likely to be detrimental to colonic health. American Journal of Clinical Nutrition 2011, 93 (5), 1062–1072. [DOI] [PubMed] [Google Scholar]
  • 42.Brinkworth GD; Noakes M; Clifton PM; Bird AR, Comparative effects of very low-carbohydrate, high-fat and high-carbohydrate, low-fat weight-loss diets on bowel habit and faecal short-chain fatty acids and bacterial populations. Brit J Nutr 2009, 101 (10), 1493–1502. [DOI] [PubMed] [Google Scholar]
  • 43.Nicholson JK; Holmes E; Kinross J; Burcelin R; Gibson G; Jia W; Pettersson S, Host-Gut Microbiota Metabolic Interactions. Science 2012, 336 (6086), 1262–1267. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

SI
NIHMS982422-supplement-SI.pdf (865.9KB, pdf)

RESOURCES