Intended for healthcare professionals

Clinical Review State of the Art Review

Long term outcomes of metabolic/bariatric surgery in adults

BMJ 2023; 383 doi: (Published 18 December 2023) Cite this as: BMJ 2023;383:e071027
  1. Anita P Courcoulas, professor of surgery1,
  2. Christopher R Daigle, service line medical director2,
  3. David E Arterburn, senior investigator, affiliate professor34
  1. 1Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
  2. 2Bariatric Surgery Program, Washington Permanente Medical Group, Bellevue, WA, USA
  3. 3Kaiser Permanente Washington Health Research Institute, Seattle, WA, USA
  4. 4Department of Medicine, University of Washington, Seattle, WA, USA
  1. Correspondence to: D E Arterburn David.E.Arterburn{at}


The prevalence of obesity continues to rise around the world, driving up the need for effective and durable treatments. The field of metabolic/bariatric surgery has grown rapidly in the past 25 years, with observational studies and randomized controlled trials investigating a broad range of long term outcomes. Metabolic/bariatric surgery results in durable and significant weight loss and improvements in comorbid conditions, including type 2 diabetes. Observational studies show that metabolic/bariatric surgery is associated with a lower incidence of cardiovascular events, cancer, and death. Weight regain is a risk in a fraction of patients, and an association exists between metabolic/bariatric surgery and an increased risk of developing substance and alcohol use disorders, suicidal ideation/attempts, and accidental death. Patients need lifelong follow-up to help to reduce the risk of these complications and other nutritional deficiencies. Different surgical procedures have important differences in risks and benefits, and a clear need exists for more long term research about less invasive and emerging procedures. Recent guidelines for the treatment of obesity and metabolic conditions have been updated to reflect this growth in knowledge, with an expansion of eligibility criteria, particularly people with type 2 diabetes and a body mass index between 30.0 and 34.9.


Obesity is a well established risk factor for developing many chronic diseases, such as type 2 diabetes, cardiovascular disease, and cancer, as well as an increased risk of covid-19 related hospital admission and death. Despite these known risks, many patients, doctors, and health policy makers remain uncertain about the long term efficacy and safety of available treatments for obesity. Recent advances in drug therapy have increased the demand for obesity treatments, but long term (five years or more of follow-up) data on drug therapy for obesity have remained relatively scant. On the other hand, long term evidence on the efficacy and safety of metabolic/bariatric surgery (MBS) has continued to accrue over the past 25 years, particularly for adults with type 2 diabetes. Recent data highlight the durability of weight loss and improvements in comorbidity, and the comparative effectiveness of alternative approaches to MBS, but they also clearly indicate risks, including the potential for weight regain, reoperation, and a higher incidence of substance use disorders and suicide. Given the important trade-offs between the long term risks and benefits of MBS, doctors should engage patients with severe obesity in a shared decision making conversation about the role of MBS in the management of their weight and related health problems. This review summarizes recent and emerging evidence related to the safety, efficacy, and metabolic outcomes of MBS to help guide clinical decision making. We also alert clinicians to emerging trends in treatment and major unanswered research questions to help to guide conversations with patients.

Sources and selection criteria

We based this review on articles found by searching PubMed and the Cochrane Library from the time of our last review (2014)1 to December 2022, with the terms “bariatric surgery”, “gastric bypass”, “sleeve gastrectomy”, “one anastomosis gastric bypass”, and “biliopancreatic diversion”. Our search was limited to English language articles. We gave priority to evidence obtained from systematic literature reviews, meta-analyses, and randomized controlled trials (RCTs) when possible. We sought to include studies that were at least one year and preferably five years or more in duration where possible. We excluded studies of MBS that did not include a control or comparator group, where possible.


The prevalence of obesity (commonly defined as body mass index (BMI) ≥30) has continued to increase, with 42.4% of adults in the US having obesity in 2017-18 and 9.2% of adults having severe obesity (BMI ≥40).2 In England in 2017, 27.4% of adults had BMI ≥30 and 2.5% of adults had BMI ≥40.3 In a report on the prevalence of obesity in 27 EU member states and Kuwait during 2018-20, 16.3% of adults had BMI ≥30 and rates of BMI ≥40 varied widely from 1.7% (Spain) to 5.5% (Kuwait).45

In the US, 198 000 bariatric operations were performed in 2020, with 61.6% being sleeve gastrectomy, 20.7% Roux-en-y gastric bypass (RYGB), 1.2% adjustable gastric banding (AGB), 0.7% one anastomosis gastric bypass (OAGB), and 11.1% revisional operations.6 Outside North America, more than 375 000 bariatric operations were performed from 2014 to 2018, including approximately 38% RYGB, 46% sleeve gastrectomy, 8% OAGB, and 5% AGB.7 The popularity of AGB has declined rapidly worldwide in recent years, whereas sleeve gastrectomy has been rising rapidly and OAGB is increasingly performed.

All MBS procedures are now routinely carried out with a less invasive, laparoscopic approach. Figure 1 shows the four most common bariatric procedures. Sleeve gastrectomy is a stomach-only operation that is performed by mobilizing the greater curvature of the stomach from its attachments and then dividing the stomach vertically around a calibrated bougie that is 36 to 40 French in size. The transected part of the stomach, which consists of approximately two thirds to three quarters of the stomach, is removed, and the new remaining stomach is a long and narrow curved tube. RYGB is a stomach and small intestine operation that consists of a small divided proximal gastric pouch and a modest small intestinal bypass of approximately 100-150 cm in length. The common channel of small intestine beyond the bypass allows for adequate absorption of nutrients. The residual stomach is not removed but is a conduit for digestive secretions only. Biliopancreatic diversion with or without duodenal switch is a gastric sleeve stomach resection with a longer intestinal bypass component at the end of the stomach. Finally, AGB is a silicone ring that is wrapped around the upper stomach, just below the esophagus, with an inflatable inner balloon to adjust the amount of gastric restriction via an infusion port placed outside the abdominal cavity. Other less common and emerging procedures are reviewed later in this article.

Fig 1
Fig 1

Common metabolic/bariatric surgery procedures. A: sleeve gastrectomy; B: Roux-en-Y gastric bypass; C: biliopancreatic diversion; D: adjustable gastric banding. Adapted from American Society for Metabolic and Bariatric Surgery. Bariatric Surgery Procedures (

Randomized controlled trials comparing bariatric surgery with medical and lifestyle therapy

In the past 15 years, 13 RCTs have compared MBS with lifestyle and medical therapy for the treatment of type 2 diabetes (table 1), showing that MBS results in significantly larger short to mid-term improvements in glycemic control, disease remission, cardiovascular risk factors, and chronic kidney disease. Findings from randomized trials of MBS resulted in international diabetes organizations publishing a joint statement supporting the consideration of MBS as a treatment option for patients with a BMI of 30.0-34.9 and type 2 diabetes.21

Table 1

Randomized controlled trials of metabolic/bariatric surgery versus medical and lifestyle treatment for type 2 diabetes treatment

View this table:

The trial with the longest follow-up to date was from Italy and randomized 60 patients to medical therapy, RYGB, or biliopancreatic diversion.20 At 10 years, type 2 diabetes remission rates were 5.5% (95% confidence interval 1.0% to 25.7%) for medical therapy (one participant went into remission after crossover to surgery), 50.0% (29.9% to 70.1%) for biliopancreatic diversion, and 25.0% (11.2% to 46.9%) for RYGB.20 People in the RYGB and biliopancreatic diversion groups had fewer diabetes related complications than those in the medical group (relative risk 0.07 (95% confidence interval 0.01 to 0.48) for both MBS comparisons). Serious adverse events occurred more commonly among participants in the surgical groups. In the US, the STAMPEDE study published five year results of a three arm trial comparing intensive medical therapy with either RYGB or sleeve gastrectomy plus intensive medical therapy. Among 150 patients in that study with type 2 diabetes and a BMI of 27-43, MBS was more effective than intensive medical therapy alone. The primary endpoint of glycated hemoglobin ≤6.0% with or without the use of diabetes medications was met by two (5%) of 38 patients who had intensive medical therapy, compared with 14 (29%) of 49 patients who underwent RYGB (adjusted P=0.03; P=0.08 in the intention-to-treat analysis) and 11 (23%) of 47 patients who underwent sleeve gastrectomy (adjusted P=0.07; P=0.17 in the intention-to-treat analysis).17 Patients who underwent the surgical procedures had a greater mean absolute reduction in glycated hemoglobin from baseline than did patients who had medical therapy alone (2.1% v 0.3%; P=0.003). The two surgical groups also had superior weight loss, better lipid concentrations, reduced insulin use, and higher quality of life scores.17

The STAMPEDE study and 12 other randomized trials either included or were specifically designed to study a population with BMI <35 (class 1 obesity), and these results also show significant improvements in type 2 diabetes in the lower BMI group (table 1). Four of these randomized trials merged to form the Alliance of Randomized Trials of Medicine versus Metabolic Surgery in Type 2 Diabetes (ARMMS-T2D) consortium that is collecting pooled, longitudinal follow-up data for participants previously randomly assigned to surgical versus non-surgical treatment of type 2 diabetes. This group of 316 participants now represents the largest cohort of patients (approximately one third with BMI 30-34.9) ever randomly assigned to surgical versus non-surgical treatments for type 2 diabetes. Three year results from ARMMS have been published and seven to 10 year results are anticipated.16 At three years, remission of diabetes was achieved in more participants after surgery than with medical/lifestyle intervention (60/160 (37.5%) v 2/76 (2.6%); P<0.001). Reductions in glycated hemoglobin and BMI were also greater after surgery.16

These RCTs have their limitations, including small sample sizes and, for most studies, inadequate duration to detect differences in the incidence of cardiovascular and end organ complications of type 2 diabetes. In addition, the definition of remission of diabetes varied between studies, as did the proportion of people with BMI <35 (table 1). Finally, the type of non-surgical treatments (lifestyle, drugs, exercise) and adherence to the program varied between studies.

Steatotic liver disease

Steatotic liver disease is a new overarching term, now used to describe the various etiologies of steatosis/fatty liver disease. Non-alcoholic fatty liver disease is now labeled as metabolic dysfunction associated steatotic liver disease (MASLD) and includes patients who have hepatic steatosis and have at least one of five cardiometabolic risk factors. A new category, termed MetALD, now describes people with MASLD who consume greater amounts of alcohol per week, and metabolic dysfunction associated steatohepatitis (MASH) is the replacement term for non-alcoholic steatohepatitis.22 Observational studies suggest that MBS may improve MASH. In a large, single center study, among people with MASH and obesity, MBS compared with non-surgical management was associated with a significantly lower risk of incident major adverse liver outcomes and major adverse cardiovascular events (MACE). Patients with MASH (n=1158: 650 who underwent MBS and 508 in a non-surgical control group) were followed for seven years. The cumulative incidence of major adverse liver outcomes at 10 years was 2.3% (0% to 4.6%) in the MBS group and 9.6% (6.1% to 12.9%) in the non-surgical group (adjusted absolute risk difference 12.4%, 95% confidence interval 5.7% to 19.7%; adjusted hazard ratio 0.12, 95% confidence interval 0.02 to 0.63; P=0.01). The cumulative incidence of MACE at 10 years was 8.5% (5.5% to 11.4%) in the bariatric surgery group and 15.7% (11.3% to 19.8%) in the non-surgical group (adjusted absolute risk difference 13.9%, 5.9% to 21.9%; adjusted hazard ratio 0.30, 0.12 to 0.72; P=0.007).23 The efficacy of MBS on MASH has recently been compared with lifestyle interventions and medical therapy in a randomized trial in Italy in which 288 people with obesity and MASH were assigned to lifestyle modification plus best medical care, RYGB, or sleeve gastrectomy. Histological resolution of MASH without worsening of fibrosis at one year follow-up was significantly higher in the RYGB (54; 56%) and sleeve gastrectomy group (55; 57%) compared with lifestyle modification (15; 16%) (P<0.001). MASH resolution was 3.60 (95% confidence interval 2.19 to 5.92; P<0.001) times greater in the RYGB group and 3.67 (2.23 to 6.02; P<0.001) times greater in the sleeve gastrectomy group compared with the lifestyle group.24 This trial complements the observational data and indicates that MBS may be more effective than lifestyle intervention for the treatment of MASH.

Cardiovascular disease, microvascular disease, and mortality

Cardiovascular disease is a leading cause of death among adults, particularly those with severe obesity. Although behavioral and pharmacological weight loss interventions can improve cardiovascular risk factors (for example, blood pressure and glycemic control) among adults with obesity, no studies have shown that non-surgical weight loss can reduce the incidence of major cardiovascular disease events.25 On the other hand, we identified 10 observational studies involving nine separate cohorts with a total of >120 000 patients who had MBS (table 2) that consistently show a significant association between MBS and a lower risk of primary or secondary cardiovascular disease events compared with non-surgical interventions or usual medical care, including several studies among adults with type 2 diabetes. A major limitation in this area is a lack of RCTs (owing to the high cost of conducting trials powered for cardiovascular disease endpoints); however, the magnitude of the effect sizes is so large in these observational studies that unmeasured confounders are unlikely to be driving this association.3536 Also, these studies have not compared MBS against newer non-surgical interventions, such as sodium-glucose cotransporter-2 inhibitors or glucagon-like peptide-1 receptor agonists, which have been shown to reduce the risk of cardiovascular events. Finally, up until 2021, most of the data have come from patients undergoing RYGB; however, two recent large observational studies suggest that sleeve gastrectomy may also be associated with better cardiovascular disease outcomes than usual non-surgical treatment.3334

Table 2

Observational studies of metabolic/bariatric surgery (MBS) and cardiovascular disease (CVD) outcomes

View this table:

Microvascular disease—As with cardiovascular disease outcomes, more evidence has emerged in recent years indicating that MBS may reduce the risk of microvascular complications of type 2 diabetes, such as nephropathy, retinopathy, and neuropathy. This includes limited data from three randomized trials comparing MBS with intensive medical-lifestyle treatment of type 2 diabetes and obesity,172037 which suggest that MBS may result in less microalbuminuria and greater improvement in estimated glomerular filtration rate (table 3). We identified additional support for improvements in microvascular outcomes after MBS from two recent systematic reviews. The first identified two RCTs and 12 observational studies involving more than 110 000 patients receiving MBS that together estimated an 83% lower relative risk (0.17, 0.13 to 0.22) of diabetic retinopathy at a median of two years’ follow-up compared with non-surgical treatment.38 The second involved three RCTs and seven observational studies including 3459 patients receiving MBS, which together suggested a 74% lower risk (odds ratio 0.26, 95% confidence interval 0.15 to 0.42) of developing any microvascular disease (composite of nephropathy, retinopathy, and neuropathy).39 In both these systematic reviews, most of the data come from the retrospective observational studies.

Table 3

Systematic reviews and randomized trials of metabolic/bariatric surgery (MBS) and microvascular disease outcomes

View this table:

Mortality—Although numerous RCTs of MBS have been conducted, none has yet been powered to investigate its effect on long term mortality compared with non-surgical treatment. However, at least 32 observational studies have examined the effects of MBS on mortality (supplementary table25) involving more than 173 000 patients receiving MBS, with a median relative reduction in mortality of 46% (range 16-89%).25

Long term comparative outcomes of gastric bypass versus sleeve gastrectomy

Sleeve gastrectomy is now the most common bariatric procedure performed worldwide, and since our last review we identified eight new randomized trials with five years or longer follow-up examining differences in outcomes between RYGB and sleeve gastrectomy.174142434445 Some studies were designed to directly compare outcomes between the procedures, whereas others prospectively compared metabolic procedures head to head with intensive medical management.1741424346 In the STAMPEDE trial, involving 49 patients having RYGB and 47 having sleeve gastrectomy, RYGB showed better mean weight loss compared with sleeve gastrectomy (−23.2 (standard deviation 9.6) kg v −18.6 (7.5) kg; P=0.01), but no statistically significant differences in diabetes outcomes were seen.1747 In the SM-BOSS trial, which compared 110 RYGB and 107 sleeve gastrectomy procedures, the authors observed no difference between RYGB and sleeve gastrectomy at five years with respect to weight loss, glycemic control, or complications requiring intervention. Of note, remission of acid reflux was observed in 60.4% of patients after RYGB compared with 25.0% after sleeve gastrectomy (P=0.002). Increasing acid reflux symptoms or escalation in reflux treatment also occurred more often after sleeve gastrectomy (31.8%) than after RYGB (6.3%) (P=0.006).41 The SLEEVEPASS five year trial compared outcomes of 119 RYGB and 121 sleeve gastrectomy procedures and showed no significant differences in weight loss, remission of type 2 diabetes, or complications at five years43; however at 10 years’ follow-up the percentage total weight loss was 3.5% higher for RYGB than sleeve gastrectomy (26.9% v 23.4%; P<0.001), but with no difference in type 2 diabetes remission rates.46 An analysis that merged five year data from the SLEEVEPASS and SM-BOSS found that percentage total weight loss was 3.2% (95% confidence interval 1.6% to 4.7%) larger with RYGB than with sleeve gastrectomy, with no difference in type 2 diabetes remission between the two procedures and more complications in the RYGB cohort.42 These findings are consistent with the multicenter US PCORnet Bariatric Study, an observational comparative effectiveness study of 24 982 patients undergoing RYGB and 18 961 patients undergoing sleeve gastrectomy, which showed significantly greater percentage total weight loss with RYGB than sleeve gastrectomy at five years (mean difference 6.7%, 5.8% to 7.7%),48 a 10% higher rate of type 2 diabetes remission with RYGB than sleeve gastrectomy (hazard ratio 1.10, 1.04 to 1.16), a 25% lower type 2 diabetes relapse rate with RYGB than sleeve gastrectomy (hazard ratio 0.75, 0.67 to 0.84),49 and a significantly lower risk of operation or intervention after sleeve gastrectomy than RYGB (hazard ratio 0.72, 0.65 to 0.79).50

When comparing long term outcomes in bariatric surgery, considering observed differences in reoperation/reintervention rates and the durability of weight loss achieved with the various surgical approaches is important. In the SM-BOSS trial, reoperation or reintervention was reported in 15.8% (16/101) of patients after sleeve gastrectomy and 22.1% (23/104) after RYGB. The most frequent indications for reintervention were acid reflux for sleeve gastrectomy and internal hernia for RYGB.41 In the 10 year SLEEVEPASS paper, the authors reported reoperation rates of 15.7% and 18.5% for sleeve gastrectomy and RYGB, respectively.46 When considering differences in weight regain between the procedures, the PCORnet Bariatric Study found that weight regain to within 5% of the preoperative baseline occurred least often among patients who had RYGB (3.3%), followed by those who had sleeve gastrectomy (12.5%) and AGB (36.0%), at five year follow-up. These findings have been corroborated by two other smaller retrospective observational studies.5152 Collectively, these studies highlight the important trade-offs between the benefits and risks of RYGB and sleeve gastrectomy, to help to inform shared decision making conversations with patients (table 4).

Table 4

Evidence summary comparing outcomes of Roux-en-y gastric bypass and sleeve gastrectomy to inform shared decision making

View this table:

OAGB is surgical procedure that uses a long gastric pouch connected by a single wide gastro-jejunal anastomosis to a loop of jejunum 150-200 cm distal to the ligament of Treitz, thus creating a gastric bypass by way of a loop and with a single anastomotic connection (fig 2).6566 In a meta-analysis of 25 RCTs comparing OAGB and RYGB, including a total of 2715 patients, RYGB showed a better weight loss after three months (two studies, 131 patients; mean difference 2.41%, 0.46% to 4.36%; I2=76%, P=0.02), six months (two studies, 69 patients; 3.83%, 2.46% to 5.21%; I2=5%, P<0.001), one year (three studies, 180 patients; 6.35%, 4.69% to 8.01%; I2=0%, P<0.001), and five years (two studies, 128 patients; 3.90%, 1.21% to 6.59%; I2=0%, P=0.005).67 In terms of remission of type 2 diabetes, two RCTs have been published but no meta-analysis has been done.6668 In the first study, among 33 patients randomized only three had type 2 diabetes—one in the OAGB group and two in the RYGB group—all of whom had remission at five years.68 In a French study, 253 patients were randomized to RYGB or OAGB, but no significant difference was found between the type 2 diabetes remission rates at two years.66 An International Federation for the Surgery of Obesity (IFSO) position statement on OAGB reviewed 95 studies with a total of 23 341 patients and found limited data on complications reported from seven out of 95 studies of OAGB. Perioperative complications occurred in 5.5% and reoperation in 1.0%, and perioperative mortality was low (<0.05%). Late complications occurred in 5.5%, including marginal ulcers, bowel obstruction, protein malnutrition, and biliary reflux, with a reoperation rate of 1.3%.65

Fig 2
Fig 2

One anastomosis gastric bypass. Adapted from ObesityGoAway. Mini gastric bypass or one anastomosis gastric bypass(OAGB) (

Observational studies have compared longer term effects of gastric bypass and sleeve gastrectomy with respect to established cardiovascular risk factors, such as hypertension and dyslipidemia.535469 In the ENGAGE CVD study, patients from nine practices having RYGB and sleeve gastrectomy (n=4964) were compared over five years to assess remission and relapse rates for hypertension. After five years, without accounting for relapse, 42% of patients having RYGB and 43% of those having sleeve gastrectomy experienced remission of hypertension. When accounting for relapse, 17% of patients having RYGB and 18% of those having sleeve gastrectomy remained in remission at five years, with no statistical differences noted between procedures.53 Another study similarly compared the effect of RYGB and sleeve gastrectomy on dyslipidemia with up to four years of follow-up. Without accounting for relapse, remission of dyslipidemia was achieved in 58.9% of patients after RYGB and 51.9% following sleeve gastrectomy after four years.54 After accounting for relapse, remission of dyslipidemia was still significantly higher after RYGB (38.0%) compared with sleeve gastrectomy (28.0%), after four years. Regarding changes in predicted cardiovascular risk, a study using the ENGAGE CVD data found no significant difference between RYGB and sleeve gastrectomy in changes in predicted 10 year risk of atherosclerotic cardiovascular disease five years after surgery.69

Risks of metabolic/bariatric surgery

The perioperative risks of MBS have declined in the laparoscopic era.70 Perioperative mortality is between 0.1% and 1.1%, and perioperative morbidity varies widely between 2% and 20% depending on both the specific type of procedure and characteristics of patients.7172 In the longer term, from a large, national comparative outcomes study of MBS procedures at five years, operation or intervention, endoscopy, and hospital admission were more likely after RYGB than after sleeve gastrectomy, but no difference in mortality was seen.50

In the past decade, more studies have assessed the potential for non-operative adverse outcomes following MBS, including the risk of substance and alcohol use disorders and suicide or accidental deaths.73 Overall, 18 observational studies with sample sizes ranging from 50 patients to >4000 patients, indicate that MBS is associated with an increased risk of alcohol and substance use disorders compared with usual care.7475 A meta-analysis of five observational studies at three years after surgery found that the pooled odds of alcohol use disorder were 1.83 (1.53 to 2.18; P<0.001) for RYGB compared with non-surgical treatment. In a matched controlled, multisite study of US veterans (predominantly men), eight years after a sleeve gastrectomy, the probability of unhealthy alcohol use was higher in surgical versus control patients (7.9% (95% confidence interval 6.4% to 9.5%) versus 4.5% (4.1% to 4.9%); difference 3.4% (1.8% to 5.0%)). Similarly, eight years after an RYGB, the probability of unhealthy alcohol use was higher in surgical than control patients (9.2% (8.0% to 10.3%) versus 4.4% (4.1% to 4.6%); difference 4.8% (3.6% to 5.9%)).62 Some mechanisms have been proposed to explain these findings, including pharmacokinetic studies showing higher peak blood alcohol concentrations after RYGB compared with controls, changes in reward sensitivity via a neurobiological mechanism, changes to the ghrelin system, and altered genetic expression in some regions of the brain.76 Together, these findings strongly suggest that education, screening, evaluation, and referral for treatment should be incorporated into both preoperative and postoperative bariatric surgical care as well as into careful lifelong monitoring in primary care settings.

An early study in the US showed that suicide rates in one state over 10 years among post-MBS patients were 13.7 per 10 000 among men and 5.2 per 10 000 among women, rates much higher than in age and sex matched US controls (2.4/10 000 men aged 35-64; 0.7/10 000 women aged 35-64).77 A systematic review including 28 studies estimated an overall suicide rate of 4.1 per 10 000 person years, which was higher than in the general population.78 Not included in this review, a more recent Canadian study examined self-harm emergencies both before and after MBS and found that the rate of these events increased from 2.3 events per 1000 people three years before surgery to 3.6 events per 1000 three years after surgery, with the most common cause being medication overdose.79 A recent multisite study of US veterans involving more than 3800 patients having MBS and 34 000 carefully matched (including for mental health conditions) non-surgical controls with a mean follow-up of 4.6 years found that the risk of suicidal ideation was still significantly higher for post-MBS patients (adjusted hazard ratio 1.21, 1.03 to 1.41), as was risk of suicide attempt (1.62, 1.22 to 2.15).64 On the other hand, another study of 12 000 cases of MBS from Western Australia showed no increased incidence of suicide or self-harm in the MBS population after an average of three years’ follow up.80 Finally, a recent Swedish registry study of more than 22 500 people showed that an increased risk of self-harm diagnoses, hospital admissions for depression, and completed suicides two years after MBS was completely attributed to a previous history of self-harm or depression that was present before the MBS procedure.81 These authors note that bariatric patients may be a particularly vulnerable population that could benefit from preoperative screening and recognition of these problems before surgery. In addition, the US American Society for Metabolic and Bariatric Surgery (ASMBS) issued a position statement on preoperative optimization before surgery that includes guidance to identify, treat, and optimize any existing psychiatric symptoms.82 These and other data in the literature clearly show signs of increased risk around harmful behaviors in patients after MBS, and the question remaining is how to best identify who, how, and why this occurs.

Metabolic/bariatric surgery and cancer

Obesity is associated with an increased risk of developing certain cancers.83 Observational data have been published with respect to cancer (risk) outcomes after MBS, mostly showing a reduction in both obesity related and all cause cancer cases and cancer related mortality. A recent systematic review identified eight observational studies including more than 600 000 patients and found that MBS was associated with a reduced risk of all types of cancer (pooled odds ratio 0.72, 0.59 to 0.87) and of obesity associated cancer (0.55, 0.31 to 0.96).84 A single large, multisite cohort study that compared 22 198 patients with severe obesity who underwent MBS and 66 427 non-surgical controls showed a 33% lower risk of incident cancer of any type (hazard ratio 0.67, 0.60 to 0.74; 488 incident cases in MBS group over 87 071 person years versus 2055 incident cases in non-surgical group over 228 010 person years) and a larger reduction in obesity associated cancers, such as postmenopausal breast cancer (0.58, 0.44 to 0.77) and endometrial cancer (0.50, 0.37 to 0.67).85 In a recently published retrospective cohort study of more than 30 000 patients, MBS (RYGB and sleeve gastrectomy) was significantly associated with a lower risk of obesity associated cancer (adjusted hazard ratio 0.68, 0.53 to 0.87) and cancer related mortality (0.52, 0.31 to 0.88).86 However, for incident colorectal cancer, some studies have reported an increased risk with MBS, whereas others report a decreased risk. In a systematic review of 18 studies and more than 12 million patients, MBS was found to be significantly protective for colorectal cancer incidence (hazard ratio 0.81; P=0.0142). The protective effect persisted for subgroups of women (relative risk 0.54; P=0.0014) but not for men (0.74; P=0.2798). No differences were found between surgical procedures.87

Obesity may accelerate and promote cancer growth by multiple potential mechanisms, including increased circulating estrogens, inflammatory cytokines, circulating adipokines, insulin, and insulin growth factor; higher cell proliferation; changes in microbiota; and epigenetic changes. Better elucidation of the specific biological mechanism(s) of effect in humans that are responsible for the observed changes in cancer risk with MBS is needed, as is identification of valid biomarkers to predict these clinical outcomes. The main concern with drawing definitive conclusions from any of the observational studies is the potential role of unmeasured confounding and selection bias, and a large, randomized trial is needed to further validate this association.

Obesity and metabolic/bariatric surgery in the era of covid-19

Several established risk factors for poor outcomes from covid-19 exist, perhaps none more initially apparent than obesity. At the outset of the covid-19 pandemic, data began to emerge that identified obesity as a predictor for greater need for hospital admission, intensive care, and mechanical ventilation, as well as mortality, after covid-19.8889 Obesity is associated with severe infection from respiratory viruses in general, and other obesity associated health conditions, such as type 2 diabetes and heart disease, can also increase this risk. Other hypotheses have included the effect of obesity on pulmonary function and perfusion, immune response, endocrine function, thromboembolic risk, and propagation by the pro-inflammatory obesity state of the hyperinflammatory response observed in severe covid-19.90

As the pandemic evolved, observational data began to suggest a protective effect of previous MBS in people with covid-19.9192939495 A retrospective cohort study compared patients with obesity admitted with covid-19 who had had previous MBS (n=2607) versus those who had not (n=122 092). Patients with previous surgery had lower mortality (7.8% v 11.2%; P<0.0001) and intubation rates (18.5% v 23.6%; P<0.001) while in hospital compared with patients who had not had MBS.91 A single site cohort study compared data from patients with previous MBS (n=11 809) matched 1:3 to patients with obesity without previous MBS. Test positivity rates were similar between the two groups; however, MBS was associated with significantly lower need for hospital admission and supplemental oxygen and less severe disease course of covid-19.92 Three systematic reviews and meta-analyses have also been published.939495 The largest of these included nine articles and 1 130 341 people. It reported significantly reduced hospital admission (odds ratio 0.44, 0.45 to 0.61), need for intensive care (0.44, 0.29 to 0.67), and mortality (0.42, 0.25 to 0.70).93

The covid-19 pandemic has highlighted another potential beneficial effect of MBS and introduced a debate about where MBS fits in the spectrum of medical necessity—should MBS be considered “essential” or “non-emergent” care? In addition, what has emerged from the pandemic are several useful consensus statements and practical delivery of care models for the surgical treatment of obesity during periods of resource strain or future pandemics.2296979899 Consistent across the consensus documents is an overarching opinion that patients with obesity and related conditions suffer when lifesaving therapies are delayed. In times of societal crisis, these patients should be prioritized for the full spectrum of obesity management therapies when possible; however, this must be balanced against local resource constraints and variables such as the perioperative cardiopulmonary risk of individuals when crisis related conditions are poor.2296979899

Shared decision making

The decision to undergo MBS should involve a shared decision making (SDM) process that considers the risks, benefits, and uncertainties of the operation. Ideally, this process should include clear communication from the surgeon, careful consideration of the patient’s values and preferences, and use of a decision aid for patients that provides objective information about all relevant treatment options and promotes involvement of patients in the decision making process. Many barriers to a high quality shared decision making process around MBS exist.100 Recently, two US health systems identified six barriers as the most important to tackle: lack of insurance coverage, lack of patients’ and providers’ understanding of insurance coverage, lack of organizational prioritization of SDM, lack of knowledge about MBS, lack of interdepartmental clarity between primary care and specialty care around who should do what in terms of SDM, and limited training on SDM conversations and tools.

SDM around MBS generally occurs in two parts: a conversation with primary care or a non-surgical medical specialist about options for managing obesity and obesity related diseases to help patients to decide whether they are interested in a referral to learn more about their options; and a conversation with a metabolic/bariatric surgeon and other members of a multidisciplinary bariatric team to discuss the benefits and risks of alternative approaches in greater detail and ensure alignment of these treatments with the patient’s preferences.101 A typical SDM workflow for MBS includes conducting a comprehensive assessment of the patient’s medical history, current health status, and overall health and wellbeing to determine whether the patient is a candidate for MBS; sharing educational materials and resources, such as information about the surgery and recovery process, potential side effects, and long term risks and benefits, to help the patient to better understand the treatment options and make a more informed decision; discussion of options including their long term risks, benefits, and uncertainties (table 4 summarizes these for the two most common operations) and tailoring of information to the individual patient’s demographic (for example, race/ethnicity102) and clinical situation using risk calculators,103104105 where possible; and clarification of the patient’s goals and expectations for treatments, as well as their own values and preferences around the risks and benefits of the different approaches.

To date, only one randomized trial has evaluated a decision aid for patients considering MBS,106 involving 152 patients. In this study, a video based decision aid seemed to result in a greater improvement than an educational booklet in terms of patients’ knowledge, decisional conflict, and outcome expectancies, but it did not have a significant effect on patients’ decisions to undergo surgery. Several other decision aids have been developed, but few have been rigorously tested.107108109110 A recent systematic review found that the quality of online information for patients about MBS is generally poor and does not meet international standards for decision making. Other more rigorously developed decision aids generally support the one randomized trial in showing that decisional conflict can be decreased and use of SDM with a decision aid could influence the choice of operation. Other risk calculators have been developed to help to provide patient specific estimates of the benefits and risks of MBS.103104111

Emerging bariatric procedures/treatments

A new field of endoscopic bariatric procedures (EBPs) has evolved in the past several years for use in people who fail to lose weight with lifestyle or drug therapies and for patients not eligible for or not interested in an MBS procedure (fig 3). The purported advantages are that these are less invasive, less costly, reversible, and repeatable. EBPs have been developed that are either gastric interventions or small bowel interventions.112113 Intragastric balloons function by taking up space in the stomach, creating a sense of fullness and delaying gastric emptying. They are generally indicated for a lower BMI threshold of 30-40 and result in modest, temporary weight loss. Several different manufacturers and types of intragastric balloons exist—single or multiple, filled with saline or gas, placed by endoscopy or swallowed in a capsule—some of which have a recommended four to 12 month limited implantation time and can be used as repeated treatments. Seven RCTs in more than 1000 people show that intragastric balloons result in significant weight reduction compared with control groups, with 10-15% total body weight loss at the time of removal of the balloon (six months) and 6-8% at 12 months.114115 The most common complications include nausea, vomiting, and abdominal pain in more than 20% of people, especially in the first weeks after placement when adaptation to the device is taking place. More serious complications such as obstruction, perforation, or death can occur; these are rare, but patients must be monitored closely for signs of balloon deflation or obstruction. In 2017 the US Food and Drug Administration (FDA) issued a warning about two newly recognized complications of fluid filled balloons—balloon hyperinflation and balloon related pancreatitis—for which patients should be carefully monitored and early removal of the device is indicated in an FDA letter.116

Fig 3
Fig 3

Emerging endoluminal metabolic/bariatric procedures. A: intragastric balloon (adapted from Intragastric balloon - Mayo Clinic); B: transpyloric shuttle (adapted from Transpyloric shuttle (TPS) (; C: primary obesity surgery endoluminal (adapted from POSE (; D: endoscopic sleeve gastroplasty (adapted from ESG. The pressure is on! Endoscopic sleeve gastroplasty versus laparoscopic sleeve gastrectomy: toward better patient allocation beyond pygmalionism - Gastrointestinal Endoscopy (; E: duodenal liner (adapted from Duodenal liner Various Novel and Emerging Technologies in Endoscopic Bariatric and Metabolic Treatments (

A transpyloric shuttle is a device that consists of a large silicone ball that is connected to a smaller one via a tether so that when food is ingested peristalsis moves the larger ball on to the pylorus, causing a temporary obstruction that delays gastric emptying. The transpyloric shuttle typically dwells in the stomach for 12 months. In two studies, weight loss results were similar to those with intragastric balloons.117118 In the pivotal randomized trial for US approval, 270 patients were randomized and the transpyloric shuttle group had 9.3% initial weight loss compared with 2.8% in sham controls at 12 months; 67% of patients with a transpyloric shuttle achieved at least 5% initial weight loss at 12 months.117 Primary obesity surgery endoluminal (POSE) uses an endoscopic approach to create full thickness gastric tissue plications (folds) along the fundus and distal body of the stomach. In one 12 month observational study of 147 patients and in two RCTs of 43 and 332 patients each, total weight loss ranged from 5% to 15% for the POSE device group.119 Endoscopic sleeve gastroplasty (ESG) is a minimally invasive endoscopic procedure that uses a full thickness suturing system to create a sleeve shaped stomach. No randomized trials have been conducted, but case series have shown weight loss of 17-19% at 12 months and improvements in type 2 diabetes and hypertriglyceridemia.112120 A meta-analysis of 1772 patients showed weight loss of 15% at six months, 16% at 12 months, and 17% at 24 months, with a severe adverse event rate of 2%.121 Compared with surgical sleeve gastrectomy in a 137 patient case matched study, ESG led to significantly less weight loss but fewer adverse events.122

Small bowel interventions include a duodenal-jejunal bypass liner that functions as a malabsorptive device preventing contact of food substances with the intestinal mucosa. It anchors from the duodenal bulb and extends to the proximal jejunum and dwells there for 12 months. Longer term results are lacking, and six month case series data show modest weight loss (6-7 kg in 19 people with BMI 34-38) and conflicting results on diabetes outcomes.123 A randomized trial including more than 300 people in the US was stopped owing to higher than anticipated rates of hepatic abscess. Duodenal mucosal resurfacing is another technique that uses a catheter to heat ablate the duodenal mucosa, which may result in improved glycemic control. A small case series study involving 39 patients showed glycated hemoglobin to be improved by 1.4% or 0.7% at six months depending on long or short ablation.124 Interestingly, no correlation was seen between magnitude of weight loss and glycemic improvement. An international multicenter case series study examined duodenal mucosal resurfacing in 46 patients with type 2 diabetes with and without fatty liver disease and showed a 0.9% reduction in glycated hemoglobin and improvement in liver enzymes at 12 months.125126 For all these EBPs, additional studies are needed on long term efficacy, outcomes of metabolic diseases, and comparisons with best available drug therapy and bariatric surgical options.


Considering the substantial advances in the understanding of the disease of obesity, its management, and MBS specifically, the leaderships of the ASMBS and IFSO recently produced a joint statement summarizing the literature and updating the recommended indications for MBS.127 These include consideration of MBS in people with lower BMIs (BMI 30-35 with metabolic disease (including type 2 diabetes) if non-surgical therapy is ineffective; BMI 35-40 with or without comorbid conditions), in people with different ethnic backgrounds (Asians with BMI >27.5), in younger and older patients, and as a bridge to other medically necessary procedures such as joint replacement or organ transplantation. The recommendation for metabolic disease treatment with MBS at lower BMI is consistent with and supported by a previous statement from international diabetes organizations from the second Diabetes Surgery Summit (DSS-II) in 2016, which summarized the large body of data from RCTs in people with class 1 obesity and type 2 diabetes.21

These guidelines now go beyond those from the National Institute for Health and Care Excellence (NICE) in the UK, which state that bariatric surgery is a recommended treatment option for people with a BMI of 40 or more or between 35 and 40 with other significant disease (for example, type 2 diabetes or high blood pressure) that could be improved if they lost weight.128 However, the NICE and ASMBS/IFSO guidelines are consistent for adults with recent onset of type 2 diabetes whose diagnosis has been made within the previous 10 years, as NICE recommends offering an expedited assessment for bariatric surgery to people with a BMI of 35 or over, for those with a BMI of 30 to 34.9, and for people of South Asian, Chinese, other Asian, Middle Eastern, Black African, or African-Caribbean family background with a BMI of 27.5 and over. Although the US Preventive Services Task Force recommends that clinicians offer or refer adults with a BMI of 30 or higher for intensive, multicomponent behavioral interventions (B recommendation), it does not provide any recommendations on MBS.129


The escalating rates of obesity and type 2 diabetes have increased interest in MBS to reduce the long term complications of these diseases. The evidence base for MBS has grown substantially over recent decades, but further research is urgently needed to fill major gaps and inform clinical decision making. Recent five year and 10 year randomized trials provide strong evidence supporting the long term effectiveness and safety of sleeve gastrectomy, RYGB, and biliopancreatic diversion. MBS has a lasting positive impact on weight loss and comorbidity improvement, especially in patients with type 2 diabetes. However, different procedures have trade-offs, with RYGB and biliopancreatic diversion offering the greatest improvement but carrying higher risks.

A major challenge to the field is the lack of a large, long term trial comparing MBS with best available medical treatment for obesity, particularly given the influx of new anti-obesity drugs, which now rival some MBS operations in terms of weight loss outcomes. We believe that the long term path forward for many patients with severe obesity will not be an either/or choice with regard to MBS and anti-obesity drugs but will be more of an adjuvant therapeutic approach. Finally, given the potential for some negative medical and psychological consequences of MBS, long term follow-up and psychosocial support are crucial. Primary care providers can help to mitigate potential harms of MBS by encouraging appropriate follow-up and screenings (box 1).130

Box 1

Recommended postoperative screening and follow-up

  • Monitor progress with weight loss and weight regain; consider evaluation by bariatric medicine provider, nutrition/medications if regain occurs

  • Avoid non-steroidal anti-inflammatory drugs

  • Consider gout and gallstone prophylaxis

  • Screen annually for depression and alcohol and substance use disorders; refer for treatment as needed

  • Encourage long term daily bariatric formulation vitamin supplementation to reduce risk of nutritional deficiencies

  • Annual nutritional monitoring, 24 hour urine calcium excretion (for biliopancreatic diversion), vitamin B12, folic acid, iron studies, 25-hydroxy vitamin D, intact parathyroid hormone

  • Bone density screening (dual x ray absorptiometry) every two years


Glossary of abbreviations

  • AGB—adjustable gastric banding

  • ASMBS—American Society for Metabolic and Bariatric Surgery

  • BMI—body mass index

  • EBPs—endoscopic bariatric procedures

  • ESG—endoscopic sleeve gastroplasty

  • FDA—Food and Drug Administration

  • IFSO—International Federation for the Surgery of Obesity

  • MACE—major adverse cardiovascular events

  • MASH—metabolic dysfunction associated steatohepatitis

  • MASLD—metabolic dysfunction associated steatotic liver disease

  • MBS—metabolic/bariatric surgery

  • NICE—National Institute for Health and Care Excellence

  • OAGB—one anastomosis gastric bypass

  • POSE—primary obesity surgery endoluminal

  • RCT—randomized controlled trial

  • RYGB—Roux-en-y gastric bypass

  • SDM—shared decision making

Questions for future research

  • Is metabolic/bariatric surgery more effective than emerging drug therapy for long term treatment of obesity and type 2 diabetes?

  • Is the long term (≥10 year) effect of sleeve gastrectomy on weight loss and comorbidity improvement as durable as that of gastric bypass?

  • What is the relative effect of sleeve gastrectomy and gastric bypass on long term rates of reoperation, hospital admission, and healthcare cost?

  • What patient level factors can predict success (weight loss, health improvements, and cost savings) after bariatric surgical procedures?

  • What are the mechanisms of effect for the increase in self-harming behaviors and the reduction in cancer risk after metabolic/bariatric surgery?

  • What clinical role do emerging metabolic/bariatric procedures have relative to the two most common, sleeve gastrectomy and gastric bypass? How do they compare for long term efficacy and safety?

How patients were involved in creation of this article

After email communication from the authors, one of our long time bariatric patient research partners agreed to review and provide feedback on our manuscript outline, plan, and final drafts


  • Series explanation: State of the Art Reviews are commissioned on the basis of their relevance to academics and specialists in the US and internationally. For this reason they are written predominantly by US authors

  • Contributors: All authors contributed equally to the planning, conduct, and reporting of the work described in the article and meet all four ICJME criteria for authorship. APC and DEA are the guarantors.

  • Competing interests: We have read and understood the BMJ policy on declaration of interests and declare the following interests: APC has received a research grant from Allurion Technologies.

  • Provenance and peer review: Commissioned; externally peer reviewed.