Gut Microbiota Serves a Predictable Outcome of Short-Term Low-Carbohydrate Diet (LCD) Intervention for Patients with Obesity

ABSTRACT To date, much progress has been made in dietary therapy for obese patients. A low-carbohydrate diet (LCD) has reached a revival in its clinical use during the past decade with undefined mechanisms and debatable efficacy. The gut microbiota has been suggested to promote energy harvesting. Here, we propose that the gut microbiota contributes to the inconsistent outcome under an LCD. To test this hypothesis, patients with obesity or patients who were overweight were randomly assigned to a normal diet (ND) or an LCD group with ad libitum energy intake for 12 weeks. Using matched sampling, the microbiome profile at baseline and end stage was examined. The relative abundance of butyrate-producing bacteria, including Porphyromonadaceae Parabacteroides and Ruminococcaceae Oscillospira, was markedly increased after LCD intervention for 12 weeks. Moreover, within the LCD group, participants with a higher relative abundance of Bacteroidaceae Bacteroides at baseline exhibited a better response to LCD intervention and achieved greater weight loss outcomes. Nevertheless, the adoption of an artificial neural network (ANN)-based prediction model greatly surpasses a general linear model in predicting weight loss outcomes after LCD intervention. Therefore, the gut microbiota served as a positive outcome predictor and has the potential to predict weight loss outcomes after short-term LCD intervention. Gut microbiota may help to guide the clinical application of short-term LCD intervention to develop effective weight loss strategies. (This study has been registered at the China Clinical Trial Registry under approval no. ChiCTR1800015156). IMPORTANCE Obesity and its related complications pose a serious threat to human health. Short-term low-carbohydrate diet (LCD) intervention without calorie restriction has a significant weight loss effect for overweight/obese people. Furthermore, the relative abundance of Bacteroidaceae Bacteroides is a positive outcome predictor of individual weight loss after short-term LCD intervention. Moreover, leveraging on these distinct gut microbial structures at baseline, we have established a prediction model based on the artificial neural network (ANN) algorithm that could be used to estimate weight loss potential before each clinical trial (with Chinese patent number 2021104655623). This will help to guide the clinical application of short-term LCD intervention to improve weight loss strategies.

Zhang et al aim to identify if differences in gut microbiota composition account for the variable outcomes of LCD in a group 51 overweight/obese Chinese patients. They find that LCD (10-25% calories from carbohydrates according to NLA guidelines) for 12 weeks resulted in greater changes in BMI and other adiposity indices than normal diet associated with reduced overall energy intake. This was independent of any major changes in fecal microbiota. The authors then show using a machine learning algorithm that 3 distinct gut microbiota species that produce butyrate are increased in the LCD group after intervention. Additionally, they find that Bacterioidaceae Bacteroides abundance at baseline is higher in the distinct weight loss LCD subgroup compared with the moderate weight loss subgroup and that this had greater predictive power using ANN analysis. This is an excellent study that is very well presented and executed and provides significant new insight into why the efficacy of LCD varies between people. My only real concern is the difference in BW at baseline between groups which needs to be mentioned in the discussion as a limitation. There is also a little ambiguity about which class of bacteria could be contributing more to the outcomes of LCD. Is it the increase in SCFA-producing bacteria during intervention or is it the high levels of Bacterioidaceae Bacteroides at baseline or is it both? Maybe a schematic diagram could help clarify this issue. Further from this point, while the role of SCFA including butyrate in regulating energy balance is discussed, what factors produced by Bacterioidaceae Bacteroides could contribute to the outcome of LCD? Finally, if the authors cannot measure SCFA in fecal samples, this needs to be mentioned as another limitation. I have the following minor comments/suggestions: Introduction, line 70: As causes of obesity, I would suggest saying psychosocial factors instead of depression and anxiety and also add genetic and epigenetic factors in there too. Introduction, line 86: Instead of "hunt for" I suggest "identify" instead. Introduction, line 96: Here I would suggest finishing the sentence after citing (8, 9) and then start a new sentence to the effect, "As a result, there is a lack of consensus as to what dietary type is superior to produce weight loss (8, 10)". Introduction, lines 106-111: I would suggest removing this section as the introduction is already very long. Introduction, line 117: I would suggest using another term other than forgotten dark matter when referring to the gut microbiota. Introduction, line 133: These studies only suggest that gut microbiota play role in obesity pathogenesis and not outcome of LCD intervention so please remove the latter statement. The authors either need to find studies on the role of gut microbiota on weight loss after LCD or after another weight loss intervention such as bariatric surgery. Results, line 179: It would help if the authors provide a statement on the reliability of three-day 24hr dietary recall. Results, line 232: Please provide citation on the previous use of the algorithm. Results, line 250: Does the change in abundances in each of these three identified bacteria to be increased after LCD correlate with weight loss? Results, line 256: It would help if the authors clarified at this stage how the two subgroups were defined for each group. Was this based on median values? Results, line 317: Please provide citation for use of ANN. Results, Figure 5: Please add in the caption title "after LCD" and also clarify in the caption itself which group of patients was analyzed.
Reviewer #2 (Comments for the Author): The manuscript by Zhang, et. al., describes the effects of a short-term carbohydrate restricted diet on patients with obesity, with a specific focus on understanding the effects of GI microbial compositional. The manuscript is generally well written and easy to follow. However, I have a few suggestions and questions.
1. The major conclusion that the relative abundance of Bacteroidaceae Bacteroides is a positive predictor of outcomes is very interesting.
Given the limitations of 16s data for accurately predicting bacterial abundances (genomic copy number of the 16S gene varies between species) the authors should consider performing qPCR for Bacteroides to determine if absolute abundance confirms the 16s sequencing data.
2. More information of the statistical methods should be included in the methods, results and figure legends.
Specifically, many different comparisons are made for each data set but no indication of how or if correction for multiple comparisons is provided. In addition, with the repeated sampling there is no mentioned for how you accounted for repeated measures.
For example, in figure 2, its indicated that several bacterial species are more abundant in LCD group post treatment via unpaired, two-sided students t-test. If you only want to compare the baseline to the end of study for each group independently at the vary least this needs to be a paired t-test with correction for multiple comparisons, as the data is a repeated measure and more than one bacterial genera is being evaluated. However, given you actually have two groups with two sampling times, a repeated measures 2-way ANOVA, with corrections for multiple comparisons is more appropriate.
This holds for all of the other analysis in the figures.
3. Minor point, but the introduction is really long. This can easily be shortened.

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address: • Point-by-point responses to the issues raised by the reviewers in a file named "Response to Reviewers," NOT IN YOUR COVER LETTER. • Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file. • Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file. Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If you would like to submit an image for consideration as the Featured Image for an issue, please contact Spectrum staff.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publicat ion Fees, including supplemental material costs, please visit our website.
Corresponding authors may join or renew ASM membership to obtain discounts on publication fees. Need to upgrade your membership level? Please contact Customer Service at Service@asmusa.org.
Thank you for submitting your paper to Microbiology Spectrum. for the variable outcomes of LCD in a group 51 overweight/obese Chinese patients. They find that LCD (10-25% calories from carbohydrates according to NLA guidelines) for 12 weeks resulted in greater changes in BMI and other adiposity indices than normal diet associated with reduced overall energy intake.
This was independent of any major changes in fecal microbiota. The authors then show using a machine learning algorithm that 3 distinct gut microbiota species that produce butyrate are increased in the LCD group after intervention.
Additionally, they find that Bacterioidaceae Bacteroides abundance at baseline is higher in the distinct weight loss LCD subgroup compared with the moderate weight loss subgroup and that this had greater predictive power using ANN analysis. This is an excellent study that is very well presented and executed and provides significant new insight into why the efficacy of LCD varies between people. My only real concern is the difference in BW at baseline between groups which needs to be mentioned in the discussion as a limitation.

Response:
We thank Reviewer 1 for your assessment on our manuscript. Yes, indeed, participants for RCT study were randomly selected and assigned into both groups. Upon completion of the study, we noticed the average body weight for LCD group was heavier than that for ND group at the baseline (BMI for ND and LCD: 28.61±2.04 vs. 30.44±3.38 kg m-2), which is a disadvantage of this study, but is difficult to avoid. This discrepancy has been discussed in the section of limitations. Response: Thank you very much for your comments. We believe there are fundamental issues with our interpretations as we are classifying SCFA-producing bacteria either between ND and LCD or between LCD_MG and LCD_DG to be analyzed and discussed. In figure 2D-F, we took advantage of robust statistical analysis, e.g. random forest to identify that Porphyromonadaceae Parabacteroides, Odoribacteraceae Butyricimonas, and Ruminococcaceae Oscillospira are critical bacteria members between ND and LCD group after intervention. They were not significantly different at baseline, but robustly increased after LCD intervention.
Based on the existing literature, these bacteria are all linked to the production of beneficial SCFAs, such as butyrate in the GI track. Meanwhile, butyrate also could stimulate the production of gut hormones (e.g., glucagon-like peptide-1, GLP-1) and decrease food intake to alleviate obesity (1) . We assume participants in LCD group could benefit from these bacteria and their metabolites upon low carbohydrate diet intervention. These findings will provide fundamental basis for our upcoming clinical studies of low carbohydrate and/or probiotics intervention.
In the subgroups of LCD intervention between LCD_MG and LCD_DG, these participants are further divided into moderate and distinct weight loss group based on their weight loss efficacy. Same as above, random forest is utilized to select the critical bacteria between two subgroups. Bacteroidaceae Bacteroides is the top listed candidate with highest relative abundance (Figure4G-I). Although previous studies demonstrated that Bacteroidetes is the largest propionate producers in the human gut(2, 3) and its level correlates with fecal levels of SCFAs (4)

Response:
We thank Reviewer 2 for your effort to assess our work. We believe that lacking of qPCR to determine the absolute abundance of Bacteroides could be a fundamental issue. However, we ran out of fecal samples collected for this LCD-based weight loss clinical intervention. This was a limitation of this study, that has been discussed in the limitation section. (Page 22, line 498-514). Regarding the paired analysis before and after LCD intervention, we actually want to compare the difference of relative abundance of each key microbiota. We agree with your comments that a repeated measure 2-way is more appropriate than the unpaired student t-test. For this aim, we performed two-way ANOVA with repeated measurement followed by a Tukey post hoc test for Figure 2F, Figure 4I and Figure   S5.

More information
As a result, all analyses except for Figure 2F were kept consistent. We repeated 2-ways ANOVA for Figure 2F. The relative abundance of Odoribacteraceae Butyricimonas was higher at end stage, but the P value＞0.05. Figure 2F has been updated in the revised manuscript. We also made corresponding change on manuscript "More specifically, the relative abundance of Ruminococcaceae Oscillospira was higher comparing to the baseline. Meanwhile, the relative abundance of Odoribacteraceae Butyricimonas had an increasing trend but did not reach statistical difference after 12-week LCD intervention. Other than these, another bacterial biomarker was identified Porphyromonadaceae Parabacteroides also had higher relative abundance after 12 weeks of LCD intervention" (Page11, line231-237). Therefore, we deleted sentences related to Odoribacteraceae Butyricimonas in discussion sections.
The statistical methods had been updated in "Other Statistical Analysis" of "MATERIALS AND METHODS" section as well as figure legends.
3. Minor point, but the introduction is really long. This can easily be shortened.
Response: Thank you very much for the suggestion. We have revised the introduction section into three pages. Thank you for responding to the Reviewer's queries. To meet the Journal's requirements on data availability, please deposit the sequencing data in one of the listed repositories: https://journals.asm.org/list-data-repositories and note this in the manuscript file.
Thank you for submitting your manuscript to Microbiology Spectrum. As you will see your paper is very close to acceptance. Please modify the manuscript along the lines I have recommended. As these revisions are quite minor, I expect that you should be able to turn in the revised paper in less than 30 days, if not sooner. If your manuscript was reviewed, you will find the reviewers' comments below.
When submitting the revised version of your paper, please provide (1) point-by-point responses to the issues I raised in your cover letter, and (2) a PDF file that indicates the changes from the original submission (by highlighting or underlining the changes) as file type "Marked Up Manuscript -For Review Only". Please use this link to submit your revised manuscript. Detailed information on submitting your revised paper are below.

Link Not Available
Thank you for the privilege of reviewing your work. Below you will find instructions from the Microbiology Spectrum editorial office and comments generated during the review.
The ASM Journals program strives for constant improvement in our submission and publication process. Please tell us how we can improve your experience by taking this quick Author Survey. Sincerely,

Steven Frese
Editor, Microbiology Spectrum Reviewer comments:

Preparing Revision Guidelines
To submit your modified manuscript, log onto the eJP submission site at https://spectrum.msubmit.net/cgi-bin/main.plex. Go to Author Tasks and click the appropriate manuscript title to begin the revision process. The information that you entered when you first submitted the paper will be displayed. Please update the information as necessary. Here are a few examples of required updates that authors must address: • point-by-point responses to the issues I raised in your cover letter • Upload a compare copy of the manuscript (without figures) as a "Marked-Up Manuscript" file. • Each figure must be uploaded as a separate file, and any multipanel figures must be assembled into one file. Please return the manuscript within 60 days; if you cannot complete the modification within this time period, please contact me. If you do not wish to modify the manuscript and prefer to submit it to another journal, please notify me of your decision immediately so that the manuscript may be formally withdrawn from consideration by Microbiology Spectrum.
If you would like to submit an image for consideration as the Featured Image for an issue, please contact Spectrum staff.
If your manuscript is accepted for publication, you will be contacted separately about payment when the proofs are issued; please follow the instructions in that e-mail. Arrangements for payment must be made before your article is published. For a complete list of Publicat ion Fees, including supplemental material costs, please visit our website.