Over the years, increasing obesity prevalence represents an important threat to national and global public health. Many strategies studied for weight loss have reported that low-carbohydrate diets showed beneficial effects on weight loss for obese patients. However, individual weight losses have varied widely within low-carbohydrate diet groups due to the influence of gut microbiota, which was found in the previous clinical studies of our team (ChiCTR1800015156). Numerous animal studies and clinical trials have evaluated the impact of probiotics on glucose and lipid metabolism as well as weight-loss interventions. However, the effect of the different probiotic strains may play a different role in the same species. Therefore, our team isolated new strains (Lactobacillus reuteri PLBK1，GDMCC No: 60828、Lactobacillus reuteri PLBK2，GDMCC No: 60829、Lactobacillus gasseri PLBK3，GDMCC No: 60830、Lactobacillus acidophilus PLBK4，GDMCC No: 60831、Bifidobacterium lactis PLBK5，GDMCC No: 60832) from the oral cavity, feces or environment of the Chinese race in the previous study. The mixed solid beverage composed of these five strains has local characteristics and meets the requirements of the national standard for food additives, which has obtained the national food production license (SC10632028100205).In our previous animal experiments, the treatment of compound preparation of the above five strains improve insulin resistance and dyslipidemia effectively in obese mice induced by high-fat diet，which obtained significant weight loss . Nevertheless, clinical trials are needed to be carried out to confirm its clinical effect. Therefore, we aimed to conduct a randomized, double-blind, placebo-controlled prospective study on the basis of previous research. In order to evaluate the efficacy and safety, as well as the effect on glucose and lipid metabolism of low-carbohydrate diet combined with probiotics in obese participants, the study will last for 12 weeks. At the same time, Magnetic Resonance Imaging (MRI) will be used to clarify change in body fat distribution, which may provide more clinical evidence for the application of this program in the prevention and treatment of obesity in the future.

This study is a randomized, double-blind, placebo-controlled clinical trial. According to the preliminary experimental results, the decrease in weight of the experimental group intervening by low-carbohydrate diet combined with probiotics compound preparation was up to (6.5±1.1) kg over the whole study period compared to that in control group adopted low-carbohydrate diet combined with placebo, which was about (4.1±0.8) kg. When taking superiority cutoff (δ) of 1.5kg, in α = 0.025 (one-sided) and the power (1-β) = 0.8, on the premise that the two groups of research objects are equal in number, the sample size of the experimental group and the control group is 19 cases respectively by using the PASS11.0 software to calculate (N1=N2=19 cases). Considering that a dropout rate of no more than 10% may occur during follow-up, this study included a total sample size of no less than 44.

Forty-four obese participants who meet the inclusion/exclusion criteria are randomly assigned to group A and group B at a ratio of 1:1. As the trial group, group A was treated with a 12-week low-carbohydrate diet combined with probiotic compound preparations, and group B was treated as a placebo control group with a 12-week low-carbohydrate diet combined with placebo treatment. Since the probiotic compound preparation is made of Lactobacillus reuteri, Lactobacillus gasseri, Lactobacillus acidophilus, Bifidobacterium lactis, and resistant dextrin (water-soluble dietary fiber), to eliminate the effect of resistant dextrin (water-soluble meal fiber) on the experimental results, the placebo is prepared from resistant dextrin (water-soluble meal fiber) and is packaged to be consistent with the appearance of the probiotic compound preparation.

During the trial period, participants will attend face-to-face visits at 4-week (visit 3), 8-week (visit 4), and 12-week (visit 5). During each follow-up visit, participants will be asked about their dietary structure, average daily calorie intake, exercise volume, adverse events, and so on. Research assistants will record their heart rate, blood pressure, height, weight, BMI, waist circumference, and waist-to-hip ratio. During the 12-week (visit 5), research assistants will record fasting blood glucose, fasting insulin, insulin resistance index (HOMA-IR), biomarkers of lipids metabolism (TG/TC/LDL/HDL), subcutaneous fat and visceral fat volume, the fat fraction (FF). Fecal samples will be collected by participants at baseline as well as at the end stage and be divided into 10 parts, each one is about 200-300mg. All fecal samples will be stored in a -80 ℃ freezer by researchers. At the end of the experiment, fecal samples will be uniformly sent to Metabo-Profile Biotechnology (Shanghai) Co., Ltd. to detect the composition of gut microbes.

The data of this project will be collected from the Department of Endocrinology and Metabolism, Zhujiang Hospital of Southern Medical University. Relevant information will be obtained by consulting the inspection system and asking patients. All data will be recorded independently by two specialized statisticians.

Data management strictly abides by relevant regulations, and any transfer of electronic data or subject-related data must be approved by the data management department. In case data is transmitted through non-secure electronic networks, corresponding data protection rules must be followed.

The statistical methods are as follow:

1. General analysis: In this study, if no special instructions are given, the data will be described and analyzed according to the following general principles. Statistics description: The mean, standard deviation and confidence interval will be given in measurement data, and if necessary, the minimum, maximum, P25, median and P75 will be given. When the non-parametric method is used, the median and interquartile range will be given. The frequency distribution and the corresponding percentage will be given in counting data. The frequency distribution and the corresponding percentage and average rank will be given in grade data. Basic principles: All statistical inferences use two-sided tests, the statistically significant test level is set at 0.05, and the confidence interval of the parameters is estimated to be 95% confidence interval. The comparison of counting data between groups depends on the theoretical frequency, using χ2 test, continuous correction χ2 test or Fisher's exact probability method. Two independent sample T-tests is used to compare normally distributed baseline measurement data between groups. For non-normal distribution baseline measurement data, the Wilcoxon rank sum test is used for comparison between groups. Continuous variables measured at multiple time points are analyzed using a Mixed effect model. p<0.05 is considered significant in the study. Missing data: The mixed linear model will be used to obtain the point estimation and standard deviation of the treatment effect. The Markov chain Monte Carlo method is used to fill in the missing data in the multivariate analysis. Abnormal data: Interquartile range whose observation value is greater than P75 or less than P25 more than 3 times will be judged as outlier data. In the analysis process, sensitivity analysis will be used for outlier data, that is, both retention and elimination of outlier data are analyzed. If the results are not contradictory, the data will be retained; if contradictory, it will be depended on the specific situation.
2. Basic characteristics: The mean, standard deviation, median, maximum and minimum will be calculated for quantitative data such as age, height, weight and BMI. The frequency and percentage will be listed for qualitative information such as gender.
3. Efficacy analysis: The efficacy analysis will be based on the Full Analysis Set (FAS). Analysis of the main efficacy indicators: the extent of weight loss during the trial period is a continuous variable, so that the results will be analyzed using Mixed effects model. In the model, subject factors are considered as random effects and intervention, time and mutual interaction effects are considered as fixed effects.Analysis of secondary efficacy indicators: After 12-week intervention, the continuous variables such as insulin resistance indicators, waist circumference, waist-to-hip ratio, lipids (TG/TC/LDL/HDL), body fat percentage (BF%), subcutaneous and visceral adipose tissue volume, fat fraction (FF) compared with the baseline will be analyzed by mixed effects model. Other categorical variables are analyzed using χ2 test, continuous correction χ2 test or Fisher's exact probability method.
4. Security analysis: The safety rating will be based on the safety set (SS), with descriptive statistical analysis as the main focus. The adverse events, serious adverse events and reactions occurred in the two groups will be analyzed. An adverse reaction is defined as an adverse event that has a relationship with the study drug as "definitely related or may be related or undeterminable". The laboratory results that are normal before the trial but abnormal after treatment will be described. The mean, standard deviation, median, minimum and maximum values of blood pressure, heart rate and laboratory indicators before and after the intervention in the two groups, as well as the number and rate of "normal turned abnormal" or "abnormal aggravation" of vital signs and laboratory indicators will be calculated.
5. Sensitivity analysis: Sensitivity analysis will be performed after missing data filling and abnormal data elimination.
6. Statistical software and general requirements: SPSS 24.0 statistical software is used for data entry and statistical analysis.

Inclusion Criteria:

• Male;
• 18 to 35 years old;
• BMI≥28kg/m2;
• No contraindications for MRI examination;
• No alcohol intake, no use of any probiotic-related foods or medications, antibiotics, hormones and weight-loss products in the past three months;
• No other metabolic diseases;
• Volunteering to participate in and cooperate with the research and sign the written informed consent.

Exclusion Criteria:

• Subjects suffered from Secondary obesity including hypercortisolism, hypothyroidism, insulinoma, drug-induced obesity, etc.

• Obesity-related complications that may be life-threatening or cause serious consequences in the short term have occurred.

• Any of the following cardiovascular diseases have occurred in the past：

  1. Myocardial infarction.
  2. Cerebral infarction.
  3. Cardiac or revascularization surgery such as coronary artery bypass graft surgery and percutaneous transluminal coronary angioplasty.
  4. Unstable angina pectoris;
  5. Congestive heart failure diagnosed as Class III or IV of the New York Heart Association.
  6. Transient ischemic attack or significant cerebrovascular disease.

• Patients with symptoms of gastrointestinal diseases, including intestinal obstruction, intestinal ulcer and gastrointestinal bleeding.

• A history of gastrointestinal surgery, such as bariatric surgery, banding surgery, gastrointestinal anastomosis and enterectomy.

• A state of known immunodeficiency, including but not limited to individuals who have experienced organ transplantation or acquired immunodeficiency syndrome (AIDS).

• Subjects who cannot comply with the protocol and patients with serious physical or psychological diseases that may affect the effectiveness or safety under the judgment of the investigator.

• Patients who cannot cooperate to complete the entire intervention and follow-up due to various other factors.