One-anastomosis gastric bypass vs sleeve gastrectomy for diabetes remission and weight loss: A meta-analysis

Abstract

BACKGROUND

One-anastomosis gastric bypass (OAGB) and sleeve gastrectomy (SG) are surgical procedures increasingly performed for weight loss and to achieve remission of diabetes mellitus. Literature comparing the medium-term efficacy of these two procedures is scarce. As such, a meta-analysis comparing OAGB and SG in terms of diabetes remission (DR) and percentage of excess weight loss (EWL) is warranted.

AIM

To compare OAGB and SG in terms of DR and EWL% in the medium term.

METHODS

A comprehensive literature search was conducted in PubMed/MEDLINE, Cochran Library, and Web of Science for relevant articles, from inception through April 2025, using the keywords “one-anastomosis gastric bypass”, “sleeve gastrectomy”, “mini-gastric bypass”, “diabetes remission”, “one-anastomosis”, and “excess weight loss”. Clinical trials, prospective, retrospective and case-control studies were included; cross-sectional studies, case reports, editorials, and opinions were excluded. The Cochrane Risk of Bias Tool and the Newcastle-Ottawa Scale were used to assess the quality of included studies, and RevMan version 5.4 was used for data analyses.

RESULTS

A total of 1360 articles were identified, and 35 studies were retrieved of which 32 were included in the final analysis. Three full texts were excluded as they did not include data on DR or EWL%. OAGB achieved higher DR than SG at 1 year following surgery [odds ratio (OR) = 1.77, 95% confidence interval (CI): 1.22-2.57, I² = 76%]. However, DR rates were similar at 3 years and 5 years following surgery (OR = 0.82, 95%CI: 0.61-1.10, I² = 23% and OR = 0.92, 95%CI: 0.31-2.72, I² = 75%, respectively). OAGB showed higher EWL% at 1 year (OR = 9.30, 95%CI: 6.45-12.15, I² = 91%), 3 years (OR = 10.02, 95%CI: 9.40-10.64, I² = 22%), and 5 years (OR = 11.61, 95%CI: 3.74-19.48, I² = 97%). OAGB showed higher late complications than adjustable SG. The results were not different in sub-group analysis including only clinical trials, observational studies, and removing studies including super-obese patients and studies contributing most to heterogeneity.

CONCLUSION

In the medium term, DR rates were similar between OAGB and SG; however, OAGB showed higher EWL% than SG, and late complications were higher in OAGB. Clinical trials investigating the predictors of DR and EWL% are recommended.

Key Words: One anastomosis gastric bypass; Sleeve gastrectomy; Diabetes remission; Excess weight loss; Meta-analysis

Core Tip: Obesity and diabetes are growing at an alarming rate. Bariatric surgery is an effective method for weight management and inducing diabetes remission (DR); therefore, choosing the correct type of bariatric surgery is important. Sleeve gastrectomy is the most frequently performed bariatric surgery, and one-anastomosis gastric bypass has seen an increase in popularity recently. Literature regarding the most effective bariatric surgery for weight reduction and DR is scarce. This review provides broader insights into one-anastomosis gastric bypass and sleeve gastrectomy in the medium term, including their impact on excess weight loss, DR, complications, mortality, and quality of life.

INTRODUCTION

The prevalence of obesity is rising at an alarming rate. According to the global estimates, more than half of the world’s population is expected to be obese or overweight by 2035[1]. The rising rate of obesity is paralleled by an increasing rate of type 2 diabetes (T2D) mellitus, which has a global prevalence of 10.5%[2]. Due to the global burden of diabetes mellitus and obesity, an effective treatment to address both of these serious conditions is a priority[3]. Bariatric/metabolic surgery is an effective and durable treatment for obesity despite the emergence of highly effective anti-obesity medications[4]. Bariatric surgery is recommended for adults with a body mass index (BMI) > 35 kg/m² regardless of the presence of obesity-related comorbidities. For individuals with T2D and a BMI > 30 kg/m² or those with a BMI < 35 kg/m² who have not achieved significant or lasting weight loss, bariatric surgery may also be recommended[5,6]. There are three types of bariatric surgery: Restrictive, malabsorptive, and combined procedures. Restrictive procedures reduce the capacity of the stomach, leading to decreased food intake and promoting a feeling of fullness (satiety) after eating smaller portions. Malabsorptive procedures induce malabsorption by altering gastrointestinal anatomy, resulting in calorie reduction and substantial weight loss. The combined procedures incorporate both mechanisms[7]. The malabsorptive and combined procedures are associated with higher rates of weight loss and better metabolic effects than the restrictive procedures; however, they also carry a higher risk of nutritional deficiencies[8]. Roux-en-Y gastric bypass (RYGB), one-anastomosis gastric bypass (OAGB), and sleeve gastrectomy (SG) are the most frequently performed bariatric surgeries. Ongoing research aims to identify the procedure with the fewest complications[9].

Since its introduction by Rutledge[10] in 1997, OAGB has been increasingly performed. OAGB has been found to effectively mitigate obesity comorbidities and improve patients’ quality of life with manageable side effects[10,11]. OAGB is a safe and effective bariatric surgery for weight reduction in individuals with morbid obesity. Its major benefits are weight reduction, significant remission of obesity-related comorbidities such as dyslipidemia, and a low risk of postoperative leak. OAGB has recently gained increasing popularity among surgeons and patients due to its relatively less invasive approach, easy reversibility and revision, lower morbidity, and promising metabolic outcomes compared to RYGB. However, the occurrence of malnutrition, ulcers, and bile reflux is higher after OAGB compared to SG[12,13].

SG is a restrictive procedure that is achieved by removing the majority of the stomach, specifically the greater curvature, leaving a smaller tubular stomach. This surgery leads to increased levels of glucagon-like peptide 1, a hormone that promotes insulin release, reduces glucagon secretion, and contributes to feelings of fullness (satiety), while simultaneously decreasing ghrelin, the “hunger hormone”. SG is easy to perform and has become the most commonly performed bariatric surgery worldwide since 2016[14,15]. The benefits of SG are weight loss, resolution of comorbidities (including diabetes, hypertension, obstructive sleep apnea, and metabolic-associated fatty liver disease), and improved quality of life[16]. The complications range from 4.4% to 12.8% with bleeding, leaks, and stenosis being the most common; long-term complications include weight regain, malnutrition, and reflux[17].

OAGB (known as a ‘mini gastric bypass’) is a simplified modification of the traditional gastric bypass procedure, which combines both restrictive and malabsorptive components of weight loss surgery and uses a single gastro-jejunal anastomosis. The advantages in the context of the primary metabolic approach are weight loss, resolution of obesity comorbidities, improvement in quality of life, and a low rate of perioperative complications. In addition, it has shown promising results in treating gastroesophageal reflux disease (GERD), particularly as a revisional procedure after other weight loss procedures[18-20]. Importantly, it offers benefits such as reversibility, convertibility, and revisability, and has gained popularity since 2022[14,21]. In this procedure, the stomach is divided between the antrum and body along the lesser curvature, with the division extending upwards (cephalad) towards the angle of His. The pouch is then anastomosed to the jejunal loop as an ante colic and ante-gastric loop gastrojejunostomy[10].

Meta-analyses comparing OAGB and SG in terms of diabetes remission (DR) and body weight reduction are scare, and have been limited by the small number of included studies, high heterogeneity, and lack of long-term follow-up (Magouliotis et al[4], Quan et al[22], and Wang et al[23]). The previous meta-analyses have also been limited by the inclusion of studies with overlapping data from the same hospital and authors, and published at the same time[24,25]. Therefore, an updated meta-analysis with more recent data on OAGB and SG outcomes is needed. To this end, this meta-analysis was conducted to compare the effects of OAGB and SG on complete DR and percentage of excess weight loss (EWL) in the medium term.

MATERIALS AND METHODS

Study design

This meta-analysis was conducted to assess the effects of OAGB and SG on complete DR and EWL%. The literature search was conducted in March 2025 and April 2025 with no limitation regarding the study period, strictly adhering to Preferred Reporting Items for Systematic reviews and Meta-Analyses guidelines.

Inclusion criteria

Clinical trials, prospective and retrospective studies, and case-control studies with a minimum period of 1 year and published in the English language were included. The studies compared OAGB and SG regarding complete DR and EWL% in the short and medium term.

Exclusion criteria

Cross-sectional studies, case reports, editorials, opinions, systematic reviews, study protocols without results, and studies with duration < 1 year were excluded. Studies that did not compare OAGB and SG in terms of DR and EWL% were also excluded.

Outcome measures

The outcome measures were comparisons between OAGB and SG in terms of EWL% and complete DR.

Remission of T2D

The definition of T2D remission varies significantly. Therefore, DR was defined as reported by each study: Fasting blood glucose < 126 mg/dL and hemoglobin A1c (HbA1c) < 6.5%[26]; fasting blood glucose < 100 mg/dL and HbA1c < 6%[27]; HbA1c < 6.0% for at least 1 year without anti-diabetes medications[28]; HbA1c < 6.5% without the use of oral hypoglycemic treatment or insulin[29]; fasting blood glucose < 100 mg/dL and HbA1c < 6%[30,31]; fasting blood glucose < 126 mg/dL and HbA1c < 6.5%[32]; and HbA1c < 6.5% without glycemic therapy and fasting plasma glucose < 7 mmol/L or HbA1c < 48 mmol/mol[33]. Some studies did not report or clearly define the criteria for DR.

Literature search

We searched the PubMed/MEDLINE, Cochran Library, and Web of Science databases for articles published in English with no limitation regarding the publication date. The literature search was conducted in March 2025 and April 2025. The keywords used were “one-anastomosis”, “single-anastomosis”, “gastric bypass”, “omega loop”, “sleeve gastrectomy”, “mini-gastric bypass”, “diabetes remission”, and “weight loss”. We identified 1360 articles, and 850 articles remained after the removal of duplicates. The excluded studies did not fulfill the inclusion criteria, which were the number of total DR and EWL%. Some studies measured partial remission/improvement of HbA1c levels. Studies that did not report the EWL% in mean ± SD and those that estimated the BMI pre-intervention and post-intervention were not included. Of the 35 texts retrieved, 32 were included in the final analysis (31 studies assessed complete DR, 21 assessed EWL%, and some studies assessed both outcomes). Three full-texts were excluded because they did not include data on DR or EWL% (Figure 1).

Figure 1  Comparison between one-anastomosis gastric bypass and sleeve gastrectomy regarding complete diabetes remission and excess weight loss (Preferred Reporting Items for Systematic reviews and Meta-Analyses chart).

Data extraction

The author used an Excel sheet to extract data from all of the included studies, and a standardized data extraction form was developed. The age of the participants who underwent OAGB and SG, sex distribution, BMI in each group, outcome measures including EWL% and DR, authors, country of publication, study type, and total number of events in OAGB and SG were collected (Tables 1, 2, and 3).

Table 1 Basic characteristics of patients who underwent conventional treatment and bariatric surgery, mean ± SD.

Ref.	Age (years), OAGB	Age (years), SG	Females, OAGB	Females, SG	BMI (kg/m2), OAGB	BMI (kg/m2), SG	Outcomes	Comments
Musella et al[24]	48.5 ± 8.7	49.2 ± 9.1	39.6%	27.3%	48.3 ± 9.2	48.1 ± 7.8	DR & EWL%	More females in OAGB
Abdel-Rahim et al[26]	42.9 ± 6.17	42.95 ± 7.63	65%	85%	50.88±3.99	47.77 ± 6.18	DR & EWL%	Higher BMI in OAGB
Ahmad et al[27]	47.1 ± 10.8	44.6 ± 10.9	59.1%	60.9%	43.04 ± 7.87	42.85 ± 5.02	DR	No significant differences at the baseline
Abdelshafy et al[28]	44.2 ± 4.99	43.6 ± 3.7	64.5%	64.5%	44.2 ± 4.99	43.6 ± 3.7	DR	No significant differences at the baseline
Dayan et al[29]	67.6 ± 2.8	67.6 ± 2.6	61.4%	61.4%	41.8 ± 7.8	43.3 ± 5.9	DR & EWL%	No significant differences at the baseline
Madhok et al[31]	45.0 ± 25	51 ± 30	47.4%	55.4%	67.0 ± 60	65.0 ± 60	DR & EWL%	OAGB were younger, more males
Milone et al[32]	34.9 ± 6.01	33.7 ± 5.61	62.2%	53.5%	47.3 ± 3.88	46.0 ± 4.77	DR & EWL%	More women in OAGB
Toksoy et al[33]	43.9 ± 12.1	40.5 ± 10.8	65.3%	59.3%	43.1 ± 3.9	42.2 ± 4.5	DR	OAGB were older
Alkhalifah et al[36]	33.8 ± 10.4.1	35.2 ± 10	70%	74.9%	40.4 ± 7.7	36.4 ± 7.6	DR & EWL%	Higher BMI in OAGB
Catro et al[37]	42.4 ± 11	43.5 ± 10.2	73.2%	75.9%	43.8 ± 9.2	45.2 ± 9.2	DR & EWL%	No significant differences at the baseline
Das et al[38]	47.1	46.7	84.2%	69.2%	47.1 ± 9.75	52.7 ± 7.17	DR & EWL%	Less body mass index in OAGB
Ding et al[39]	46	30	15.4%	80%	34.7	43.8	DR	OAGB were older, fewer females, and had lower BMI
Gambardella et al[40]	40.1 ± 5.1	38.9 ± 4.6	45%	44.5%	47.1 ± 2.3	46.2 ± 3.6	DR & EWL%	No significant differences at the baseline
Jammu and Sharma[41]	38	23	71.2%	45.4%	42.5	35	DR	OAGB were older, more females, and high BMI
Jain et al[42]	42.9±14.0	39.9±11.7	38.6%	35%	44.3±7.88	44.5±7.16	DR & EWL%	OAGB were older
Kansou et al[43]	41.2 ± 11.3	41.2 ± 12.3	93.4%	91.9%	42.8 ± 5.0	43.4 ± 6.5	DR & EWL%	No significant differences at the baseline
Kular et al[44]	NA	NA	NA	NA	NA	NA	DR & EWL%	NA
Moradi et al[45]	47.3 ± 10.2	47.5 ± 10.9	77.3%	80.1%	44.6 ± 7.0	43.8 ± 6.9	DR	No significant differences at the baseline
Plamper et al[46]	42. ± 11	44 ± 11	76.2%	61.8%	54.1 ± 6.6	54.6 ± 10.3	DR & EWL%	OAGB younger, more females, less BMI
Ruiz-Tovar et al[47]	43.8 ± 11.5	43.9 ± 10.9	75%	75%	45 ± 4.1	46.5 ± 3.4	DR & EWL%	No significant differences at the baseline
Schmitz et al[48]	39.11 ± 0.9	41.57 ± 1.07	74.5% vs 58.1%	58.1%	64.14 ± 0.3	66.91 ± 0.6	DR & EWL%	More females, lower BMI in OAGB
Seetharamaiah et al[49]	42.89 ± 14.02	39.89 ± 11.75	62%	65%	44.32 ± 7.88	44.57 ± 7.16	DR & EWL%	No significant differences at the baseline
Shen et al[50]	42.9	42.9	NA	NA	39.0 ± 7.2	39.0 ± 7.2	DR	NA
Shivakumar et al[51]	42.9 ± 14.02	39.9 ± 11.75	61.3%	65%	44.3 ± 7.88	44.6 ± 7.16	DR & EWL%	OAGB were older
Singla et al[52]	Matched	Matched	Matched	Matched	> 50	> 50	DR & EWL%	No significant differences at the baseline
Tabesh et al[53]	41.54 ± 11.20	41.42 ± 11.77	76%	84.4%	46 ± 6.15	455.71	24 years, 17.2% vs 12.9% diabetes	Higher BMI in OAGB
Toh et al[54]	47 ± 9	40 ± 11	69.8%	61.8%	40.3 ± 9.1	43±7.9	DR & EWL%	OAGB were older with lower BMI
Vrakopoulou et al[55]	46.6 ± 7.8	45.9 ± 7.5	60%	57.1%	52.7 ± 10.8	52.2 ± 8.6	DR & EWL%	No significant differences at the baseline
Wazir et al[56]	48.21 ± 9.77	48.21 ± 9.77	70.2% females	70.2% females	49.80 ± 6.225	49.80 ± 6.225	DR	Not assessed
Sari et al[57]	20-59	20-59	79.6% females	79.6% females	≥ 35	≥ 35	DR & EWL%	Not assessed
Yang et al[58]	NA	NA	NA	NA	32.1 ± 10.3	33.9 ± 9.4	DR & EWL%	No significant differences at the baseline

OAGB: One-anastomosis gastric bypass; NA: Not available; SG: Sleeve gastrectomy; BMI: Body mass index; DR: Diabetes remission; EWL: Excess weight loss.

Table 2 Diabetes mellitus complete remission.

Ref.	Country	Study type	OAGB, 1 year	SG, 1 year	OAGB, 3 years	SG, 3 years	OAGB, 5 years	SG, 5 years
Musella et al[24]	Italy	Retrospective	82/96	67/110
Abdel-Rahim et al[26]	Egypt	Prospective	11/20	11/20
Ahmad et al[27]	Syria	Prospective	110/137	57/92
Abdelshafy et al[28]	Egypt	Trial	18/22	14/22
Dayan et al[29]	Israel	Retrospective	36/41	25/83
Lee et al[30]	Taiwan	Trial	18/30	9/30
Madhok et al[31]	United Kingdom	Retrospective	4/6	9/17
Milone et al[32]	Italy	Prospective	14/16	10/15
Toksoy et al[33]	Turkey	Retrospective	94/108	32/39
Alkhalifah et al[36]	Taiwan	Prospective	496/533	186/205			488/533	205/205
Catro et al[37]	Spain	Retrospective	114/123	68/83			110/123	63/83
Das et al[38]	United Kingdom	Retrospective	3/4	3/5
Ding et al[39]	China	Retrospective	7/10	1/3
Gambardella et al[40]	Italy	Prospective	48/60	79/131	53/57	105/128
Jammu and Sharma[41]	India	Retrospective	59/473	13/339
Jain et al[42]	India	Trial	49/49	47/47			40/55	37/52
Kansou et al[43]	France	Retrospective	25/27	19/21
Kular et al[44]	India	Retrospective	58/63	49/61
Moradi et al[45]	Iran	Retrospective	509/675	142/201	336/675	94/201
Plamper et al[46]	Germany	Retrospective	302/3 19	83/98	NA	NA	NA	NA
Ruiz-Tovar et al[47]	Spain	Trial	66/70	53/61			61/70	50/61
Schmitz et al[48]	Germany	Retrospective	34/51	42/45
Seetharamaiah et al[49]	India	Trial	41/49	36/47
Shen et al[50]	Taiwan	Retrospective	64/81	91/130
Shivakumar et al[51]	India	Trial	41/49	36/47	46/49	44/47
Singla et al[52]	India	Retrospective	58/75	64/75
Tabesh et al[53]	Iran	Retrospective	2/35	1/151
Toh et al[54]	Singapore	Retrospective	23/32	60/73
Vrakopoulou et al[55]	Greece	Retrospective	22/25	10/28
Wazir et al[56]	United Kingdom	Retrospective	1/2	10/18
Sari et al[57]	Turkey	Retrospective					27/28	29/31

Data are events/total. NA: Not available; OAGB: One-anastomosis gastric bypass; SG: Sleeve gastrectomy.

Table 3 Excess weight loss following one-anastomosis gastric bypass and sleeve gastrectomy at 1-year post-surgery, mean ± SD/n.

Ref.	Country	Study type	OAGB, 1 year	SG, 1 year	OAGB, 3 years	SG, 3 years	OAGB, 5 years	SG, 5 years
Musella et al[24]	Italy	Retrospective	64.7 ± 22.9/96	52.4 ± 18.3/110			22.8 ± 5.9/96	20.1 ± 5.3/110
Abdel-Rahim et al[26]	Egypt	Prospective	95.11 ± 7.00/20	78.48 ± 19.07/20
Dayan et al[29]	Israel	Retrospective	67.2 ± 22.3/41	45.8 ± 18.0/83
Lee et al[30]	Taiwan	Trial					78.2 ± 19.7/519	68.7 ± 30.3/519
Madhok et al[31]	United Kingdom	Retrospective	58.0 ± 7.75/19	45 ± 21.5/56	NA	NA	NA	NA
Milone et al[32]	Italy	Prospective	24.19 ± 4.42/16	24.33 ± 4.48/15
Alkhalifah et al[36]	Taiwan	Prospective	84.5 ± 35.2/533	64.8 ± 34.3/205
Catro et al[37]	Spain	Retrospective	72.5 ± 16.4/123	68.8 ± 18.6/83
Das et al[38]	United Kingdom	Retrospective	50.2 ± 28.6/19	49.9 ± 19.5/26
Gambardella et al[40]	Italy	Prospective	83.6 ± 18.1/60	74.3 ± 13.8/131
Jain et al[42]	India	Trial	65.9 ± 10.9/101	64.8 ± 14.3/100	67.5 ± 16.6/101	61 ± 26.4/100	65.3 ± 13.9/101	55.9 ± 27/100
Kansou et al[43]	France	Retrospective	79.3 ± 17.8/136	71.4 ± 19 /136
Kular et al[44]	India	Retrospective	63 ± 21.2/104	69 ± 22.5/118	70 ± 22.6/104	61 ± 26.4/118	68 ± 24/104	51.2 ± 23/118
Plamper et al[46]	Germany	Retrospective	66.2 ± 13/169	57.3 ± 19/118	NA	NA	NA	NA
Ruiz-Tovar et al[47]	Spain	Trial					97.9 ± 7/70	76.3 ± 6/61
Schmitz et al[48]	Germany	Retrospective	36 ± 0.8/150	29 ± 1.2/93	42.5 ± 1.9	32.4 ± 2.7
Seetharamaiah et al[49]	India	Trial	66.87 ± 10.87/101	63.97 ± 13.24/100
Shivakumar et al[51]	India	Trial	66.2 ± 10.9/101	63.9 ± 13.5/100	66.5 ± 15.7/93	61.2 ± 25.2/92
Singla et al[52]	India	Retrospective	74.57 ± 13.2/75	56.20 ± 18.9/75
Toh et al[54]	Singapore	Retrospective	68 ± 28.5/40	61.2 ± 20/195	66.2 ± 35.6/31	47.9 ± 22.8/53	65.2 ± 27.5/8	47.3 ± 27.5/15
Vrakopoulou et al[55]	Greece	Retrospective	98.2 ± 29.0/115	79.7 ± 14.5/437
Sari et al[57]	Turkey	Retrospective					77.66 ± 30.55/62	71.11 ± 27.81/129
Yang et al[58]	Taiwan	Prospective	72 ± 20/89	67.2 ± 18.4/32

Data are events/total. NA: Not available; OAGB: One-anastomosis gastric bypass; SG: Sleeve gastrectomy.

Risk of bias of the included studies

The Cochrane Risk of Bias (RoB) 2 tool[34] and Newcastle Ottawa Scale were used to assess the quality of the included studies[35]. The Cochrane RoB 2 was used to evaluate the studies for potential biases across five key domains, namely selection bias, performance bias, attrition bias, detection bias, and reporting bias. Each domain was evaluated as low RoB, high RoB, and some concerns. The Newcastle Ottawa Scale was used to assess three fundamental aspects of the methodology: Selection of the participants, 0-4 points; confounder adjustment, 0-2; and the determination of outcome indicators, 0-3. A study with a score of 7-9 points was defined as high quality (Tables 4 and 5).

Table 4 Newcastle Ottawa Scale risk of bias of the included studies.

Ref.	Selection	Compatibility	Exposure	Total score
Musella et al[24]	4	2	2	8
Abdel-Rahim et al[26]	3	2	2	7
Ahmad et al[27]	3	2	3	8
Dayan et al[29]	4	2	2	9
Madhok et al[31]	3	1	2	7
Milone et al[32]	4	2	2	8
Toksoy et al[33]	4	2	3	9
Alkhalifah et al[36]	3	2	2	7
Catro et al[37]	4	2	3	9
Das et al[38]	4	2	2	8
Ding et al[39]	3	2	2	7
Gambardella et al[40]	3	2	3	8
Jammu and Sharma[41]	3	2	2	7
Kansou et al[43]	4	2	2	8
Kular et al[44]	4	2	2	8
Moradi et al[45]	4	2	3	8
Plamper et al[46]	3	2	2	7
Schmitz et al[48]	4	2	2	8
Shen et al[50]	4	2	2	7
Singla et al[52]	3	2	2	7
Tabesh et al[53]	3	2	3	8
Toh et al[54]	4	2	2	8
Vrakopoulou et al[55]	N/A	N/A	N/A	N/A
Wasir et al[56]	4	1	2	7
Sari et al[57]	3	2	2	7
Yang et al[58]	3	2	1	6

N/A: Not applicable.

Table 5 Risk of bias assessment of the included studies according to the Cochrane risk of bias tool for randomized controlled trials.

Ref.	Selection bias1	Selection bias2	Performance bias	Attrition bias	Detection bias	Reporting bias	Overall bias
Abdelshafy et al[28]	Some concern	Low	Some concern	Low	Low	Some concern	Some concerns
Lee et al[30]	Some concern	Some concern	Low	Low	Low	High	Low
Jian et al[42]	Low	Low	High	Low	Some concern	Some concern	Some concerns
Ruiz-Tovar et al[47]	Low	Low	Some concerns	Some concerns	Some concerns	Low	Some concerns
Seetharamaiah et al[49]	Low	Some concerns	Some concerns	Some concerns	Some concerns	Low	Some concerns
Shivakumar et al[51]	Low	Low	Some concerns	Some concerns	Low	Low	Low

¹Random sequence generation bias.

²Allocation concealment.

Statistical analyses

Review Manager version 5.4.1 (Cochrane Collaboration, Oxford, United Kingdom) was used for the data analyses. To analyze data from multiple studies, specifically focusing on the outcomes of complete T2D remission and EWL%, forest plots were used to visualize and summarize the data. Odds ratios (ORs) were used to compare outcomes for dichotomous data, whereas mean differences (MD) were used for continuous data. A random effects meta-analysis was used to pool studies with similar characteristics, particularly when significant heterogeneity existed, and 95% confidence intervals (CIs) were calculated to quantify the precision of the pooled effect estimate. The I² statistic was used to evaluate the degree of heterogeneity among studies. I² value < 25% was considered low heterogeneity, whereas a value > 50% indicated substantial heterogeneity, and a random effects model was used. We generated funnel plots to assess potential publication bias in meta-analysis including 10 or more studies. A sub-group analysis was conducted to assess the outcomes in clinical trials and observational studies and after excluding studies with super-obese patients to identify the source of heterogeneity. In addition, the included studies were removed one by one, and their contribution to the heterogeneity was estimated, finally, a sub-group analysis was conducted after removing studies contributing most to the heterogeneity (Tables 6 and 7). A P-value < 0.05 was considered statistically significant.

Table 6 Effects of different studies on heterogeneity (all patients with diabetes remission).

Ref.	Effect
Musella et al[24]	1% increase
Abdel-Rahim et al[26]	1% increase
Ahmad et al[27]	1% increase
Abdelshafy et al[28]	1% increase
Dayan et al[29]	9% decrease
Lee et al[30]	1% increase
Madhok et al[31]	1% increase
Milone et al[32]	1% increase
Toksoy et al[33]	1% increase
Alkhalifah et al[36]	1% increase
Catro et al[37]	1% increase
Das et al[38]	1% increase
Ding et al[39]	1% increase
Gambardella et al[40]	1% increase
Jammu and Sharma[41]	1% decrease
Jian et al[42]	No effect
Kansou et al[43]	1% increase
Kular et al[44]	1% increase
Moradi et al[45]	2% decrease
Plamper et al[46]	No effect
Ruiz-Tovar et al[47]	1% increase
Schmitz et al[48]	7% decrease
Seetharamaiah et al[49]	1% increase
Shen et al[50]	1% increase
Shivakumar et al[51]	1% increase
Singla et al[52]	3% decrease
Tabesh et al[53]	No effect
Toh et al[54]	2% decrease
Vrakopoulou et al[55]	2% decrease
Wazir et al[56]	1% increase
Sari et al[57]	1% increase

Table 7 Effects of different studies on heterogeneity (all patients with excess weight loss).

Ref.	Change
Musella et al[24]	1% increase
Abdel-Rahim et al[26]	1% increase
Dayan et al[29]	No change
Lee et al[30]	1% increase
Madhok et al[31]	4% decrease
Milone et al[32]	No change
Alkhalifah et al[36]	No change
Catro et al[37]	1% increase
Das et al[38]	1% increase
Gambardella et al[40]	1% increase
Jian et al[42]	No change
Kansou et al[43]	1% increase
Kular et al[44]	No change
Plamper et al[46]	1% increase
Schmitz et al[48]	1% increase
Seetharamaiah et al[49]	1% increase
Shivakumar et al[51]	1% increase
Singla et al[52]	No effect
Toh et al[54]	1% increase
Vrakopoulou et al[55]	No effect
Sari et al[57]	No effect
Yang et al[58]	1% increase

RESULTS

Characteristics of the included studies

A total of 32 studies[24,26-33,36-56] were included in the meta-analysis, comprising 19 from Asia, 11 from Europe, and 2 from Africa. Twenty-six studies were observational (21 retrospective studies and 4 prospective studies), whereas six were controlled trials. Regarding complete DR, 31 studies were included in the analysis with 5334 patients and 3626 events[24,26-33,36-57]. OAGB achieved higher complete DR at 1 year following surgery compared to SG (OR = 1.77, 95%CI: 1.22-2.57, χ² = 115.43, P = 0.002, SMD = 28). However, high heterogeneity was found (I² = 76%, P < 0.001; Figure 2A and B).

Figure 2 Complete diabetes remission following one-anastomosis gastric bypass and sleeve gastrectomy. A: At 1 year following surgery (forest plot); B: At 1 year following surgery (funnel plot); C: At 3 years following surgery; D: At 5 years following surgery. OAGB: One-anastomosis gastric bypass; CI: Confidence interval.

No significant difference was found regarding complete DR after 3 years of follow-up in the analysis of three studies that included 1148 patients and 678 events [OR = 0.82, 95%CI: 0.61-1.10, χ² = 2.61, P = 0.18, standard MD (SMD) = 2][40,45,51]. No significant heterogeneity was found (I² = 23%, P = 0.27; Figure 2C). No significant difference was observed regarding complete DR after 5 years of follow-up in the analysis of five studies that included 1241 patients and 1110 events[36,37,42,47,57] (OR = 0.92, 95%CI: 0.31-2.72, χ² = 16.03, P = 0.87, SMD = 4). Significant heterogeneity was found (I² = 75%, P = 0.003; Figure 2D).

EWL% was higher following OAGB compared to SG after 1 year of follow-up in the analysis of 20 studies with 4251 patients (OR = 9.30, 95%CI: 6.45-12.15, χ² = 220.41, P < 0.001, SMD = 19)[24,26,29,31,32,36-38,40,42-44,46,48,49,51,52,54,55,58]. However, high heterogeneity was found (I² = 91%, P < 0.001; Figure 3A and B). EWL% was higher following OAGB after 3 years of follow-up in the analysis of five studies with 935 patients (OR = 10.02, 95%CI: 9.40-10.64, χ² = 5.13, P < 0.001, SMD = 4)[42,44,48,51,54]. No significant heterogeneity was found (I² = 22%, P = 0.27; Figure 3C). EWL% was higher following OAGB at the 5-year follow-up in the analysis of seven studies in 2004 (OR = 11.61, 95%CI: 3.74-19.48, χ² = 191.75, P = 0.004, SMD = 6)[24,30,42,44,47,54,57]. Substantial heterogeneity was found (I² = 97%, P < 0.001; Figure 3D).

Figure 3 Excess weight loss following one-anastomosis gastric bypass and sleeve gastrectomy. A: At 1 year following surgery (forest plot); B: At 1 year following surgery (funnel plot); C: At 3 years following surgery; D: At 5 years following surgery. OAGB: One-anastomosis gastric bypass; CI: Confidence interval; OR: Odds ratio.

Sub-group analysis for complete DR: A sub-group analysis was conducted including only clinical trials in which complete DR was higher in OAGB (86.6%) compared to SG (76.8%), OR = 2.15, 95%CI: 1.31-3.54, χ² = 1.66, P = 0.003, SMD = 4. No heterogeneity was found (I² = 0%, P = 0.80; Figure 4A). Similarly DR was higher in OAGB (86.6%) compared to SG (76.8%) in observational studies, OR = 1.69, 95%CI: 1.10-2.61, χ² = 113.59, P = 0.02, SMD = 23. Significant heterogeneity was found (I² = 80%, P < 0.001; Figure 4B). DR was higher in OAGB compared to SG after removing studies with super-obese patients, OR = 2.15, 95%CI: 1.50-3.06, χ² = 66.48, P < 0.001, SMD = 21. Significant heterogeneity was found (I² = 68%, P < 0.001; Figure 4C). In a sub-group analysis of studies contributing most to heterogeneity, DR at after 1 year following surgery was not different, OR = 2.05, 95%CI: 1.26-3.36, χ² = 26.00, P = 0.004, SMD = 14. No significant heterogeneity was found (I² = 46%, P = 0.03; Figure 4D). DR was not different in OAGB and SG after five years of follow-up when including observational studies and clinical trials, OR = 1.72, 95%CI: 0.06-52.13, χ² = 17.91, P = 0.76, SMD = 2. Significant heterogeneity was found (I² = 89%, P < 0.001), and OR = 0.80, 95%CI: 0.43-1.51, χ² = 0.24, P = 0.50, SMD = 1. No significant heterogeneity was found (I² = 0%, P = 0.62; Figure 4E and F). In a sub-group analysis of studies contributing most to heterogeneity, DR at after 5 years following surgery was not different, OR = 0.59, 95%CI: 0.36-0.94, χ² = 2.56, P = 0.03, SMD = 3. No significant heterogeneity was found (I² = 0%, P = 0.46; Figure 4G).

Figure 4 Complete diabetes remission following one-anastomosis gastric bypass and sleeve gastrectomy. A: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy in clinical trials; B: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy in observational studies; C: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with super-obesity; D: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with high heterogeneity; E: At five years after one-anastomosis gastric bypass and sleeve gastrectomy in observational studies; F: At five years after one-anastomosis gastric bypass and sleeve gastrectomy in clinical trials; G: At five years after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with high heterogeneity. OAGB: One-anastomosis gastric bypass; CI: Confidence interval; OR: Odds ratio.

Sub-group analysis for EWL%: Regarding EWL%, OAGB achieved higher weight loss than SG in clinical trials and observational studies, MD = 2.13, 95%CI: 0.16-4.10, χ² = 0.54, P = 0.03, SMD = 2, no significant heterogeneity was found (I² = 0%, P = 0.76), and MD = 10.83, 95%CI: 7.35-14.31, χ² = 195.93, P < 0.001, SMD = 16. However, high heterogeneity was found (I² = 92%, P < 0.001; Figure 5A and B). EWL% was higher in OAGB when removing studies with super-obesity, MD = 6.26, 95%CI: 2.71-9.80, χ² = 86.88, P = 0.0005, SMD = 12, and significant heterogeneity was found (I² = 86%, P < 0.001; Figure 5C). In a sub-group analysis of studies contributing most to heterogeneity, EWL at after 5 year following surgery was not different, MD = 4.80, 95%CI: 2.45-7.14, χ² = 8.25, P < 0.001, SMD = 5. No significant heterogeneity was found (I² = 39%, P= 0.14; Figure 5D). EWL% was not different in OAGB and SG at 5 years in observational studies, MD = 10.83, 95%CI: 1.50-23.16, χ² = 20.06, P = 0.09, SMD = 2, and significant heterogeneity was found (I² = 90%, P < 0.001; Figure 5E). However, higher EWL was found in clinical trials, MD = 13.68, 95%CI: 4.39-22.97, χ² = 45.03, P = 0.004, SMD = 2, and significant heterogeneity was found (I² = 96%, P < 0.001; Figure 5F). In a sub-group analysis of studies contributing most to heterogeneity, EWL at after 5 year following surgery was not different, MD = 11.42, 95%CI: 6.63-16.22, χ² = 5.46, P < 0.001, SMD = 3. No significant heterogeneity was found (I² = 45%, P = 0.14; Figure 5G). Long-term complications were higher in OAGB compared to SG, OR = 1.82, 95%CI: 1.44-2.29, χ² = 2.13, P < 0.001, SMD = 5. No significant heterogeneity was found (I² = 0%, P = 0.83; Figure 5H).

Figure 5 Percentage of excess weight loss. A: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy in clinical trials; B: At 1 year after-anastomosis gastric bypass and sleeve gastrectomy in observational studies; C: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with super-obesity; D: At 1 year after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with high heterogeneity; E: At five years after one-anastomosis gastric bypass and sleeve gastrectomy in observational studies; F: At five years after one-anastomosis gastric bypass and sleeve gastrectomy in clinical trials; G: At five years after one-anastomosis gastric bypass and sleeve gastrectomy after removing studies with high heterogeneity; H: Late complications rate in one-anastomosis gastric bypass and sleeve gastrectomy. OAGB: One-anastomosis gastric bypass; CI: Confidence interval; OR: Odds ratio.

Sensitivity analysis/influence analysis

A sub-analysis, removing the studies one by one, located the sources of heterogeneity (Tables 4 and 5).

DISCUSSION

In this meta-analysis, complete DR was higher in patients who underwent OAGB compared to those who underwent SG at 1 year (OR = 1.77, 95%CI: 1.22-2.57, χ² = 15.43). However, the rates were similar between the two procedures at 3 years and 5 years following surgery (OR = 0.82, 95%CI: 0.61-1.10 and OR = 0.92, 95%CI: 0.31-2.72). Regarding the EWL%, OAGB was superior to SG at 1 year, 3 years, and 5 years following surgery (OR = 9.30, 95%CI: 6.45-12.15; OR = 10.02, 95%CI: 9.40-10.64; and OR = 11.61, 95%CI: 3.74-19.48, respectively). The results were not different in sub-group analysis in clinical trials, observational studies, and in a sub-analysis excluding studies with super-obesity. Barzin et al[59] in their meta-analysis found no differences between OAGB and SG regarding DR in the short term and medium term, with a higher rate of EWL% in OAGB. The findings were similar to our findings regarding EWL%; however, our results differed regarding DR as we found a higher rate of DR following OAGB after 1 year[59]. Plausible explanations could be the small number of included studies in the previous meta-analysis (only nine retrospective studies). In addition, Barzin et al[59] only included patients with a BMI > 50 kg/m².

Another meta-analysis by Magouliotis et al[4] found a higher EWL% in the OAGB group compared to the SG group in line with current findings. However, the authors only included 10 studies and assessed the outcomes at only 1 year following surgery. Importantly, Magouliotis et al[4] performed incorrect data extraction for EWL% from the article by Kansou et al[43] and unfortunately the number of females from Jammu and Sharma’s study[41] were inaccurately entered, leading to questions about their ultimate results. Furthermore, many recent studies have been published since the meta-analysis by Magouliotis et al[4], and they were included in the current meta-analysis. Our results are not in line with the study by Ali et al[60], who included only 10 studies and found that OAGB had superiority in terms of DR compared to SG. However, the current findings align with Ali et al[60] in terms of EWL%.

Our results were also similar to those by Wu et al[13], who found that OAGB resulted in higher EWL% compared to SG; however, our results differed in terms of DR, as we found a higher rate of DR in patients who underwent OAGB at the 1-year follow-up. The higher rate of DR in OAGB was not sustained at the 3-year and 5-year follow-up. The meta-analysis by Wu et al[13] was limited by its inclusion of studies published by the same authors as well as studies with overlapping patient populations from the same hospitals. In addition, we included more recent studies in the current meta-analysis. The difference in DR rates between OAGB (15.7%) and SG (35%) over time could be influenced by the varying rates of diabetes recurrence after initial remission[61,62]. Other plausible explanations for the discrepancies are the duodenal exclusion effect and biliopancreatic limb length effect[63,64]. More studies including randomized controlled trials but with relatively few patients have recently been published. For example, a literature review by Ding et al[65] showed better DR and EWL% in OAGB compared to SG; however, the authors only included two studies that assessed the outcomes at 5 years. Kermansaravi et al[66] included only eight trials and found similar results to the current findings regarding EWL%; however, the authors found no difference in DR between OAGB and SG at 1 year following surgery, and higher DR after OAGB at 5 years post-surgery in contrast to our findings. We found higher DR after OAGB at 1 year, but this was not sustained at 3 years and 5 years. The discrepancy in results could be explained by the differences in baseline HbA1c, BMI, and duration of diabetes mellitus[67,68].

The mechanisms of DR in bariatric surgery remain unclear; however, weight loss, dietary restriction, effects on gut hormones, bile acid alteration, and microbiota dysregulation are thought to play significant roles[69]. In the present meta-analysis, we found a higher rate of late complications in OAGB compared to SG, our findings were different to Wang et al[23] who included only three studies and found a lower rate of late complications in OAGB, the higher rates of late complications in this study could be explained by higher ulcers and malnutrition in OAGB as reported by Barzin et al[59] and Ali et al[60]. Although OAGB achieved higher weight loss and DR, the procedure is associated with higher nutritional deficiencies, particularly of vitamins and minerals, as well as anemia[70,71]. OAGB is associated with higher rates of ulceration, whereas SG is associated with higher rates of GERD, necessitating conversion surgery[38]. The low rate of GERD after OAGB is explained by the use of a wider gastric tube, leading to low intraluminal pressure[4]. On the other hand, some studies have reported a higher risk of GERD after OAGB, particularly bile reflux, potentially increasing the risk of cancer[72]. Although this risk is minimal according to the International Federation for the Surgery of Obesity and Metabolic Disorders[11], experts recommend against performing OAGB in patients with severe esophagitis (grade C or grade D) or Barrett’s esophagus[73]. One study found that OAGB is associated with less craving for fatty and sweet flavors compared to SG[74]. Another study found that the overall complication rates of OAGB and SG were not significantly different[60]. Regarding mortality, literature is scarce due to the continued reservation of some surgeons regarding OAGB[75]. Singhal et al[76] found no difference in 30-day morbidity and mortality between OAGB and SG. Some studies suggest a better quality of life after gastric bypass surgery compared to SG, potentially due to lower rates of GERD and lower BMI[77,78]. Literature suggests that bariatric surgeries generally lead to a better quality of life than non-surgical approaches, with RYGB offering more substantial and durable long-term benefits than SG. Data on the long-term cost effectiveness of OAGB are lacking[79]; however, some studies have shown that RYBG is more cost-effective than SG[80]. Although SG is more cost-effective in individuals with a BMI between 35 kg/m² and 39.9 kg/m²[81].

The choice between OAGB and SG could be based on case by case because of the existing controversy regarding the rate of complications (higher GERD in SG and higher ulcer and nutritional deficiency in OAGB). Although, OAGB achieved higher EWL% and short-term DM; however, DM was not maintained in the medium-term. Due to the above, patients could be categorized based on the degree of weight loss needed and the presence of early and late complications.

Study strengths

The strength of this study is the inclusion of recent studies that were not included in previous meta-analyses[29,33,36,40,46,57], and the exclusion of studies published by the same authors to avoid overlap[80,81]. In addition, this meta-analysis has the largest up-to-date number of studies. Finally, we provide valuable information on the long-term impact of the procedures on DR and EWL% by reporting the outcomes at 3 years and 5 years post-procedures.

Study limitations

The current results should be viewed in light of the following limitations. The majority of included studies were observational (26 studies) with only six clinical trials with only a medium-term follow-up period. Significant heterogeneity was observed (> 50%), and studies were pooled using different methodologies. No heterogeneity was found when assessing clinical trials only. However, the heterogeneity persisted in observational studies indicating methodological issues. In this meta-analysis, we could not compare OAGB and SG effects on quality of life and mortality due to insufficient studies. A major limitation of this meta-analysis is that, the study was conducted by a single author, which may have increased the RoB.

CONCLUSION

OAGB leads to a higher rate of complete DR at 1-year post-operation compared to SG; however, no significant differences were evident at 3 years and 5 years after surgery. OAGB achieves superior EWL% compared to SG at 1 year, 3 years, and 5 years following surgery. The late complications were higher in OAGB. The current findings indicated that the choice between OAGB and SG needs careful patient selection depending on basic characteristics and comorbidities. Patients’ categorization depending on the BMI and comorbidities including GERD, ulcers, and nutritional deficiencies, could help the surgeon to choose between OAGB and SG. Larger controlled trials investigating the predictors of complete DR and EWL% and type of late complications are recommended.

ACKNOWLEDGEMENTS

The author gratefully acknowledges the Saudi Digital Library for free access to databases. We would like to acknowledge Ihab Farah from the Faculty of Science, University of Tabuk, for the substantial contribution to data analysis.