ABSTRACT

The aim of this study was to monitor the impact of a preoperative low-calorie diet and bariatric surgery on the bacterial gut microbiota composition and functionality in severe obesity and to compare sleeve gastrectomy (SG) versus Roux-en-Y gastric bypass (RYGB). The study also aimed to incorporate big data analysis for the omics results and machine learning by a Lasso-based analysis to detect the potential markers for excess weight loss. Forty patients who underwent bariatric surgery were recruited (14 underwent SG, and 26 underwent RYGB). Each participant contributed 4 fecal samples (baseline, post-diet, 1 month after surgery and 3 months after surgery). The bacterial composition was determined by 16S rDNA massive sequencing using MiSeq (Illumina). Metabolic signatures associated to fecal concentrations of short-chain fatty acids, amino acids, biogenic amines, gamma-aminobutyric acid and ammonium were determined by gas and liquid chromatography. Orange 3 software was employed to correlate the variables, and a Lasso analysis was employed to predict the weight loss at the baseline samples. A correlation between Bacillota (formerly Firmicutes) abundance and excess weight was observed only for the highest body mass indexes. The low-calorie diet had little impact on composition and targeted metabolic activity. RYGB had a deeper impact on bacterial composition and putrefactive metabolism than SG, although the excess weight loss was comparable in the two groups. Significantly higher ammonium concentrations were detected in the feces of the RYGB group. We detected individual signatures of composition and functionality, rather than a gut microbiota characteristic of severe obesity, with opposing tendencies for almost all measured variables in the two surgical approaches. The gut microbiota of the baseline samples was not useful for predicting excess weight loss after the bariatric process.

摘要

本研究的目的是监测术前低热量饮食和减肥手术对严重肥胖患者肠道菌群组成和功能的影响，并比较袖状胃切除术 (SG)与Roux-en-Y 胃旁路术 (RYGB)。该研究还旨在通过基于套索的分析将组学结果的大数据分析和机器学习结合起来，以检测过度减肥的潜在标志物。招募了 40 名接受减肥手术的患者（14 名接受了 SG，26 名接受了 RYGB）。每个参与者贡献了 4 个粪便样本（基线、饮食后、手术后 1 个月和手术后 3 个月）。使用 MiSeq (Illumina) 通过 16S rDNA 大规模测序确定细菌组成。通过气相和液相色谱测定与粪便中短链脂肪酸、氨基酸、生物胺、γ-氨基丁酸和铵浓度相关的代谢特征。使用 Orange 3 软件来关联变量，并采用 Lasso 分析来预测基线样品的重量损失。仅在体重指数最高的情况下观察到 Bacillota（以前的 Firmicutes）丰度和超重之间的相关性。低热量饮食对成分和目标代谢活动几乎没有影响。RYGB 对细菌组成和腐烂代谢的影响比 SG 更深，尽管两组的过度体重减轻相当。在 RYGB 组的粪便中检测到显着更高的铵浓度。我们检测到组成和功能的个体特征，而不是严重肥胖的肠道微生物群特征，在两种手术方法中几乎所有测量变量都有相反的趋势。