Baseline characteristics and alteration of body profile after bariatric surgery. The baseline characteristics of 22 patients who provided their fecal samples before and after bariatric surgery are summarized in Table 1. The mean age was 37.8 years and 8 of the 22 patients (36.4%) were male. More than half of the patients had comorbidities, including DM, hypertension, and sleep apnea. Before bariatric surgery, the mean body weight of patients was 106.8 kg (range, 67.3 – 166.0 kg), and the mean BMI was 39.2 kg/m2 (range, 30.1 – 62.1 kg/m2). Six months after bariatric surgery, the body weight and BMI of patients decreased to an average 84.2 kg (range, 55.1 – 142.8 kg) and 30.6 kg/m2 (range, 22.3 – 45.6 kg/m2), respectively (Fig. 1A).
Table 1
Baseline characteristics of enrolled patients.
Characteristics | Total (n=22) |
Age | 37.8 (21 – 64) |
Sex | |
Male | 8 (36.4) |
Female | 14 (63.6) |
Comorbidities | |
Diabetes Mellitus | 14 (63.6) |
Hypertension | 12 (54.5) |
Depression | 1 (4.5) |
Musculoskeletal pain | 4 (18.2) |
Sleep apnea | 14 (63.6) |
Dyslipidemia | 7 (31.8) |
GERD | 2 (9.1) |
Body weight (kg) | 106.8 (67.3 – 166.0) |
BMI (kg/m2) | 39.2 (30.1 – 62.1) |
Type of surgery | |
LSG | 14 (63.6) |
LRYGB | 8 (36.4) |
Data are presented as mean (min – max) or number (percentage). |
GERD, gastroesophageal reflux disease; BMI, body mass index; LSG, laparoscopic sleeve gastrectomy; LRYGB, laparoscopic Roux-en-Y gastric bypass |
Changes in gut microbial diversity and composition before and after bariatric surgery. Compared with before bariatric surgery, alpha diversity significantly increased and beta diversity differed after surgery (Fig. 1B). After bariatric surgery, Firmicutes decreased and Bacteroidetes increased at the phylum level, and the taxonomy composition at the genus level was remarkably changed (Fig. 1C). We performed linear discriminant analysis (LDA) effect size (LEfSe) to compare the gut microbial changes at the genus level following bariatric surgery. The LDA scores were computed for features that showed differential abundance of patients before and after bariatric surgery. As shown in Fig. 2A, at the genus level Blautia, Catenibacterium, and Clostridiaceae were enriched in patients before bariatric surgery, whereas Bacteroides and Streptococcus were the preponderance in patients after bariatric surgery. In addition, we performed LDA effect size analysis at the species level and identified that Ruminococcus gnavus and Blautia obeum wexlerae were enriched in patients before bariatric surgery, whereas Bacteroides thetaiotaomicron, Bacteroides nordii, Bacteroides uniformis, and Bacteroides dorei vulgatus were the preponderance in patients after bariatric surgery (Supplementary Fig. 2A).
Correlation between a specific gut microbiome and BMI. Based on the results of LDA scores, we evaluated the correlation analysis between gut microbiota at the genus level and BMI and found that Blautia showed a positive correlation with BMI (r=0.59, p=2.3x10−5) while Bacteroides had a negative correlation with BMI (r=-0.65, p=1.8x10−6) (Supplementary Table 1, Fig. 2B and 2C). The abundance of Blautia significantly decreased and the abundance of Bacteroides significantly increased after bariatric surgery (Supplementary Fig. 1A and 1B). When converting as a log value, we identified that the log value of the Blautia/Bacteroides ratio showed a positive correlation with BMI (Fig. 2D). In addition, we found a correlation between gut microbiota at the species level and BMI; Ruminococcus gnavus and Blautia obeum wexlerae showed positive correlations with BMI, while Bacteroides thetaiotaomicron, Bacteroides nordii, Bacteroides uniformis, and Bacteroides dorei vulgatus had negative correlations with BMI (Supplementary Fig. 2B and 2C).
Comparison according to the surgery type. Before surgery, the median body weight and BMI were higher in patients who underwent LSG than those in patients who underwent LRYGB, and these values significantly decreased after both LSG and LRYGB surgery (Supplementary Fig. 3A). Alpha diversity significantly increased in patients who underwent LRYGB, but not in patients who underwent LSG. Further, beta diversity of before and after surgery was significantly differ in both surgery types (Fig. 3A). The taxonomy composition at the phylum and genus levels showed changes after both LSG and LRYGB (Supplementary Fig. 3B). LDA effect size at the genus level showed similar results regardless of the surgery type (Fig. 3B). Further, the log value of the Blautia/Bacteroides ratio showed a positive correlation with BMI in both surgery types (Fig. 3C).
Validation in severely obese patients and healthy volunteers. To validate these results, we selected 42 patients with severe obesity (BMI > 35) from the 58 patients who were initially enrolled in this study for scheduled bariatric surgery and compared them to the healthy volunteers with normal BMI (n=41). The baseline characteristics of patients with high BMI and healthy controls are summarized in Table 2. Compared to a healthy person, severely obese patients showed different beta diversity, but alpha diversity was not significantly differed (Fig. 4A). The taxonomy composition of severely obese patients was different than those of healthy volunteers; severely obese patients showed more enriched Firmicutes and deficient Bacteroidetes at the phylum level and more abundant Blautia and deficient Bacteroides at the genus level (Fig. 4B).
Table 2
Baseline characteristics between healthy volunteer with normal body weight (BMI, 18.8 – 22.8) and patients with severe obesity (BMI > 35).
Characteristics | Healthy volunteers (n=41) | Patients with severe obesity (n=42) |
Age | 35.9 (22 – 59) | 36.8 (19 – 58) |
Sex | | |
Male | 13 (31.7) | 19 (45.2) |
Female | 28 (68.3) | 23 (54.8) |
BMI | 20.9 (18.8 – 22.8) | 42.5 (35.1 – 62.1) |
Comorbidities | | |
Diabetes Mellitus | 0 (0.0) | 19 (45.2) |
Hypertension | 0 (0.0) | 17 (40.5) |
Depression | 0 (0.0) | 5 (11.9) |
Musculoskeletal pain | 0 (0.0) | 6 (14.3) |
Sleep apnea | 0 (0.0) | 36 (85.7) |
Dyslipidemia | 0 (0.0) | 13 (31.0) |
GERD | 0 (0.0) | 7 (16.7) |
Data are presented as mean (min – max) or number (percentage). |
BMI, body mass index |
The results of LDA effect size demonstrated that Blautia, Streptococcus, Ruminococcus_0, Catenibacterium, Collinsella, Eubacterium_0, Dorea, Lactobacillus and Clostridium were much more enriched in severely obese patients, whereas Bacteroides, Faecalibacterium, Ruminococcus, Enterococcus, Rhizobium, Oscillospira, Alistipes, Lactococcus, Paraprevotella, and Parabacteroides were enriched in healthy volunteers. These results were similar in the Cladogram analysis (Fig. 5A). In addition, we evaluated the correlation analysis between BMI and gut microbiota of severely obese patients and healthy volunteers, and we found that Blautia showed a positive correlation with BMI and Bacteroides had a negative correlation with BMI (Fig. 5B and 5C). We also identified that the log value of the Blautia/Bacteroides ratio showed a positive correlation with BMI (Fig. 5D).