GLP-1 medication vs bariatric surgery: what studies found

As of 8 October 2026, the studies we read that directly compare bariatric surgery with semaglutide or tirzepatide are observational, not randomized. In the observational studies we read, bariatric surgery was associated with greater weight loss than medication at 1 to 3 years, and results for death and for heart events pointed in different directions.

Wegovy injection (under the skin) and Wegovy tablets (by mouth) are FDA-approved for weight management in certain people; Ozempic injection, Ozempic tablets and Rybelsus tablets are approved only for adults with type 2 diabetes (checked September 27, 2026). Among tirzepatide products, Zepbound injection, given under the skin, is FDA-approved for weight management in certain adults, while Mounjaro injection is approved only for uses in people with type 2 diabetes (checked October 14, 2026).

Quick comparison

GLP-1 medication and bariatric surgery compared on the studies we read as of 8 October 2026; the weight-loss row takes its figures from one observational study of both treatments, and every other row comes from separate studies of each treatment or from studies that compare the two. Figures in different rows come from different groups of people and cannot be set against each other. Figures in this table are averages or percentages for groups of people, and individual results vary. Evidence levels on this page follow our grading key: Emerging means some human evidence that falls short of consistent randomized trials, such as small or single trials, indirect comparisons, or associations in observational studies, and Not graded means we give no level, here because the observational studies we read conflict. In this table, Not yet checked by us means we have not yet reviewed the evidence for that cell. Wegovy injection and Zepbound injection each carry a boxed warning about thyroid C-cell tumors seen in rodents.

Sources: Wegovy injection prescribing information, Revised 06/2026; Zepbound injection prescribing information, Revised 08/2026.

Randomized trials that compare the two

As of 8 October 2026, in searches of Europe PMC, PubMed and ClinicalTrials.gov, we found no published results from a randomized trial that compared sleeve gastrectomy or gastric bypass directly with semaglutide or tirzepatide. We found two registered randomized trials that assign people to one of these operations or to one of these drugs, and neither has posted results.

One plans 140 participants and compares surgery, semaglutide and tirzepatide, with weight loss at one year as its main measure (ClinicalTrials.gov NCT06803888).[1] The other plans 120 participants who have an advanced form of fatty liver disease, compares surgery with liraglutide, semaglutide or tirzepatide, and takes liver scarring as its main measure (ClinicalTrials.gov NCT06374875).[2]

We describe one randomized trial, published in 2013, that compared surgery with an older drug of the same class.[7] In that trial of 108 adults with obesity, type 2 diabetes and high blood pressure, diabetes was in remission at 12 months in 90% of the gastric bypass group and in none of the other two groups, one given usual care and one given usual care plus exenatide.[7] The 2013 trial result is rated Emerging: it is one trial of 108 people, and exenatide is an older drug whose results do not carry over to semaglutide or tirzepatide.

Weight loss in studies that compare the two

In a 2026 study of 44,025 adults with a BMI of 35 or higher at two New York City health systems, surgery was associated with greater weight loss than semaglutide or tirzepatide.[3] The weight-loss comparison is rated Emerging: it is an association seen in observational studies, in which people were not randomly assigned to a treatment.

Three years after sleeve gastrectomy the average loss was 22.0% of body weight, and after gastric bypass 28.4%.[3] For semaglutide at three years the paper's summary reports 7.4% among everyone prescribed it and 8.8% among people with a full year of continuous orders.[3] For tirzepatide at two years it reports 10.8% and 11.9% for the same two groups.[3] About a third of the people prescribed either drug had stopped after two monthly orders.[3] People were not randomly assigned, and the two groups differed in age, insurance and health system, so those who had surgery may differ from those who took medication in ways the study could not measure.[3] A US National Institutes of Health grant supported one of the authors.[3]

Wegovy injection (under the skin) and Wegovy tablets (by mouth) are FDA-approved for weight management in certain people; Ozempic injection, Ozempic tablets and Rybelsus tablets are approved only for adults with type 2 diabetes (checked September 27, 2026). Among tirzepatide products, Zepbound injection, given under the skin, is FDA-approved for weight management in certain adults, while Mounjaro injection is approved only for uses in people with type 2 diabetes (checked October 14, 2026).

The other observational comparisons of weight that we read as of 8 October 2026, including studies of liraglutide and of semaglutide at doses below 2.4 mg, also reported more weight loss in the surgery group; we ran no search designed to find every such study.

The authors of the 2026 New York study report that weight loss on semaglutide and tirzepatide in their study was lower than in the drugs' initial clinical trials.[3] In a separate health-record study of 7,881 adults without type 2 diabetes who started semaglutide or tirzepatide between 2021 and 2023, average weight loss at one year was 8.7% overall: 11.9% in people who did not stop the drug, 6.8% in people who stopped between 3 and 12 months, and 3.6% in people who stopped within three months.[8] In that health-record study, 80.8% of people were on low maintenance doses, and its authors say the average was lower than in the main trials, likely because of higher rates of stopping and lower maintenance doses.[8]

Some analyses compare surgery with GLP-1 drugs indirectly, by combining trials that tested each against lifestyle care or placebo.[9][10][11] We read three such analyses, and they differ on how close tirzepatide comes to surgery. A 2026 analysis of 30 trials estimated about 10 percentage points more weight loss with surgery than with GLP-1 drugs in the first two years, and for tirzepatide alone the difference from surgery was not statistically significant.[9] A second analysis, of 23 trials, concluded that tirzepatide at 10 or 15 mg and sleeve gastrectomy gave equivalent weight loss.[10] A third, of 139 trials, concluded that surgery appears superior in the long term, that semaglutide and tirzepatide showed no inferior short-term results compared with surgery, and that long-term data were lacking for most of the medicines.[11] An indirect comparison can suggest a direction but cannot show that two treatments are equal. These indirect comparisons are rated Emerging and show a direction only.

People who have bariatric surgery and people who take a GLP-1 drug differ in ways that statistical matching or weighting cannot fully remove, so an observational comparison can show an association and cannot show that the treatment caused the result.

Blood sugar and type 2 diabetes

Among the observational studies we read, a 2026 study found that HbA1c, a measure of average blood sugar, fell about as much on semaglutide or tirzepatide as after surgery, and two pooled analyses found lower HbA1c after surgery.[3][12][13] We give the blood sugar comparison between surgery and GLP-1 drugs no evidence level, because the observational studies we read conflict.

In the 6,082 people with diabetes in the 2026 study of 44,025 adults at two New York City health systems, average HbA1c fell by about as much with semaglutide or tirzepatide as after surgery over three years.[3] The authors describe the difference as not statistically significant, and they note that many patients had limited follow-up data.[3] Two pooled analyses of observational studies point the other way.[12][13] One, of 6 retrospective cohorts, found a final HbA1c lower after surgery by 1.31 percentage points.[12] The other, of 11 studies, found a smaller HbA1c reduction with GLP-1 drugs, a difference of 0.99 points.[13] Neither lists its studies in its abstract, so whether the New York study is in either pool is not known.[12][13]

A pooled follow-up of four US trials that randomly assigned adults with type 2 diabetes to surgery or to medical and lifestyle care found that HbA1c fell further after surgery: the difference between the groups was 1.4 percentage points at 7 years and 1.1 at 12 years.[4] The follow-up enrolled 262 of the 305 eligible participants; at 7 years 18.2% of the surgery group and 6.2% of the medical group were in remission, and at 12 years 12.7% and none.[4] Remission meant an HbA1c below 6.5% with no diabetes medication.[4] A quarter of the people assigned to medical care later had surgery, the surgery group included gastric banding as well as gastric bypass and sleeve gastrectomy, and use of incretin drugs, the group that includes GLP-1 drugs, rose in both groups during follow-up.[4] In the same follow-up, anemia, fractures and gastrointestinal adverse events were more common after surgery.[4] These trials compared surgery with medical care in general and were not trials against semaglutide or tirzepatide. The US National Institutes of Health funded the follow-up, one of the four original trials had a grant from a maker of surgical devices, all four received bridge funding from two such makers, and a drug maker supplied medicines during one original trial.[4] The remission finding is rated Emerging: the four trials were small and were pooled by their own investigators.

Deaths in studies that compare the two

Death from any cause after bariatric surgery and on GLP-1 drugs has been compared in observational studies that looked back at existing health records. We give the comparison of death rates no evidence level, because the observational studies we read point in different directions and differ in the people studied, the drugs and the follow-up.

The studies we read that compare death rates after surgery and on GLP-1 drugs are observational, and they do not agree. Of the 11 we read as of 8 October 2026, found by searching PubMed and Europe PMC and screening the results with a computer filter and by title rather than reading every abstract, two found surgery associated with lower mortality, seven found a GLP-1 drug or tirzepatide associated with lower mortality, and two found no significant difference. Of those 11 studies, the two that favored surgery were in adults with type 2 diabetes and obesity. One was in a US health system (3,932 adults, followed for a median of 5.9 years).[14] The other was in Israel (6,070 adults, median 6.8 years), where the association held for people who had had diabetes for 10 years or less and was not statistically significant for people with longer-standing diabetes.[15] The drugs in the Israeli study were older ones, mainly liraglutide.[15] Of those 11 studies, the seven that favored the drug included a 2025 database study of adults with a BMI of 30 or higher in which tirzepatide was associated with lower mortality than surgery.[16] The others were in people with fatty liver disease, cirrhosis, heart failure with preserved ejection fraction, hypertrophic cardiomyopathy and breast cancer, and in a semaglutide study that counted only deaths during a hospital stay for acute coronary syndrome.[17][18][19][20][21][22] Of those 11 studies, the two with no significant difference overall were in people with sleep apnea and in people with fatty liver disease and type 2 diabetes; in the second, subgroup analyses associated gastric bypass with higher mortality than semaglutide (hazard ratio 1.64, 95% CI 1.01 to 2.67) and sleeve gastrectomy with lower mortality (hazard ratio 0.44, 95% CI 0.23 to 0.85).[23][24] Of those 11, seven draw on the same records network, so they may include some of the same patients, and we read 8 as published summaries only. None of these studies can show that either treatment caused the difference.

Heart and circulation events

Heart attack, stroke, heart failure and related events after bariatric surgery and on GLP-1 drugs have been compared in observational studies that looked back at existing health records, insurance claims and registries. We give the comparison of heart and circulation events no evidence level, because the observational studies we read point in different directions.

For heart attack, stroke, heart failure and related events, the observational studies we read also point in different directions. Of the 15 we read as of 8 October 2026, found by searching PubMed and Europe PMC and screening the results with a computer filter and by title rather than reading every abstract, nine found surgery associated with fewer events,[24][25] three found a GLP-1 drug or tirzepatide associated with fewer events, two found no significant difference, and one found semaglutide or tirzepatide associated with fewer heart failure events but no significant difference in heart attack or stroke.[19] Of the 15 studies, two measured a short window: one counted heart failure only during a hospital stay for acute coronary syndrome and favored surgery, and one counted cardiac events in the 90 days after knee replacement and favored the drug.[22][26] Of those 15 studies, the nine that favored surgery include studies of adults with type 2 diabetes in Israel, Sweden and the United States, a study of adults with type 2 diabetes using nationwide registries, two US studies that compared surgery with semaglutide, and the hospital-stay study.[14][22][27][28][29][30][31] The registry study also found higher rates of alcohol or substance abuse, self-harm and fractures after surgery.[29] Of the 15 studies, the two with no significant difference were an Israeli study of older GLP-1 drugs and a study of tirzepatide in people with sleep apnea.[15][23] Of those 15 studies, the three that favored the drug were a 2025 database study of tirzepatide in adults with a BMI of 30 or higher, a study of people with hypertrophic cardiomyopathy, and the knee replacement study.[16][20][26] Of the 15 studies, the one with a split result was in people with heart failure with preserved ejection fraction.[19] None of these studies assigned treatment at random. Of the 15, six name the same records network, so they may include some of the same patients, and a seventh uses a national research network that its summary does not name.

GLP-1 medication in more detail

In US health records for 125,474 adults who started liraglutide, semaglutide or tirzepatide between 2018 and 2023, about 65% of those without type 2 diabetes and about 47% of those with it had stopped within a year.[5] These figures cover 2018 to 2023 and may not reflect current use.[5]

In that study of US health records, among people without type 2 diabetes who stopped the drug and had a weight recorded at that point, about 36% had restarted within a year.[5] The figures on stopping GLP-1 medication are rated Emerging: they describe what health records show and do not say why people stopped.

In one trial funded by semaglutide's maker (STEP 4), adults who had reached semaglutide 2.4 mg during a 20-week run-in and were then switched to placebo saw their body weight rise by 6.9% on average by week 68, while those who stayed on semaglutide lost a further 7.9%.[32] In one trial funded by the drug's maker (SURMOUNT-4), adults without diabetes who had lost weight during 36 weeks on tirzepatide and were then switched to placebo saw their body weight rise by 14.0% on average over the next 52 weeks, while those who stayed on tirzepatide lost a further 5.5%.[33] STEP 4 and SURMOUNT-4 were separate trials with different lengths of treatment before the switch to placebo, so their results cannot be compared directly.

The Wegovy and Zepbound side-effect tables come from different trials with different participants, and each label says its rates cannot be directly compared with rates from trials of another drug.[34][35]

The Wegovy label's boxed warning says that in rodents, semaglutide causes thyroid C-cell tumors, and that it is unknown whether Wegovy causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans, because the human relevance of the rodent tumors has not been determined.[34] The Wegovy label also warns about acute pancreatitis, acute gallbladder disease, low blood sugar, kidney injury from dehydration, severe stomach and bowel reactions, serious allergic reactions, diabetic eye complications in people with type 2 diabetes, heart rate increase, and inhaling stomach contents during anesthesia or deep sedation.[34] The Wegovy label reports average resting heart-rate increases of 1 to 4 beats per minute with Wegovy compared with placebo in its weight-loss trials.[34]

According to the Wegovy prescribing information, "Because clinical trials are conducted under widely varying conditions, adverse reaction rates observed in the clinical trials of a drug cannot be directly compared to rates in the clinical studies of another drug and may not reflect the rates observed in practice."[34] The Wegovy label's side-effect table for the 2.4 mg injection pools 3 placebo-controlled trials of up to 68 weeks in 2,116 adults with obesity or overweight, 19% of whom had type 2 diabetes.[34] In Table 3 of the Wegovy label, the side effects reported most often with the 2.4 mg injection were nausea (44% vs 16% on placebo), diarrhea (30% vs 16%), vomiting (24% vs 6%) and constipation (24% vs 11%).[34] In the Wegovy label's pooled weight-loss trials, 6.8% of people on the 2.4 mg injection and 3.2% on placebo stopped treatment because of side effects.[34]

The Zepbound label's boxed warning says that in rats, tirzepatide causes thyroid C-cell tumors, and that it is unknown whether Zepbound causes thyroid C-cell tumors, including medullary thyroid carcinoma (MTC), in humans, because the human relevance of the rat tumors has not been determined.[35] The Zepbound label also warns about severe stomach and bowel reactions, kidney injury from dehydration, gallbladder problems, pancreatitis, serious allergic reactions, low blood sugar, worsening diabetic eye disease in people with type 2 diabetes, and inhaling stomach contents during anesthesia or deep sedation.[35] The Zepbound label reports an average heart-rate increase of 1 to 3 beats per minute with Zepbound, compared with no increase on placebo.[35]

According to the Zepbound prescribing information, "Because clinical trials are conducted under widely varying conditions, adverse reaction rates observed in the clinical trials of a drug cannot be directly compared to rates in the clinical trials of another drug and may not reflect the rates observed in practice."[35] The Zepbound label's Table 1 pools two trials of up to 72 weeks, one of which enrolled only people with type 2 diabetes; 25% of the pooled patients had type 2 diabetes.[35] Nausea and diarrhea were the most common side effects in Table 1 of the Zepbound label: nausea was reported by 25% to 29% of people on Zepbound vs 8% on placebo, and diarrhea by 19% to 23% vs 8%.[35] In the Zepbound label's pooled weight-loss trials, 4.8%, 6.3% and 6.7% of people on Zepbound 5 mg, 10 mg and 15 mg stopped treatment because of side effects, compared with 3.4% on placebo, most often during the first few months and because of stomach and bowel side effects.[35]

Sources: Wegovy injection prescribing information, Revised 06/2026; Zepbound injection prescribing information, Revised 08/2026.

If you have severe symptoms, call 911 or seek emergency care. Report side effects to FDA MedWatch.

Bariatric surgery in more detail

In a US study of 418 people who had gastric bypass, average weight was 35% below the starting weight at 2 years, 28% below at 6 years and about 27% below at 12 years.[6] This was an observational study, funded by the US National Institute of Diabetes and Digestive and Kidney Diseases and others.[6]

In the same study, people who had sought surgery and not had it were about 2% below their starting weight at 12 years.[6]

In a 2021 analysis that pooled 58 studies and more than 3.6 million patients, 0.08% of patients died around the time of surgery (0.05% after sleeve gastrectomy and 0.09% after Roux-en-Y gastric bypass).[36] In a US registry of 611,619 patients who had surgery from 2015 to 2018, 0.8% of sleeve gastrectomy patients and 2.3% of gastric bypass patients had another operation within 30 days, and 3% and 5.9% were readmitted to hospital.[37]

A pooled analysis of 74 studies found anemia in 26% of people after gastric bypass: 15% in the first year and 35% beyond five years.[38] Studies with inadequate iron, vitamin B12 and folic acid supplementation reported higher anemia rates than studies with adequate supplementation.[38] A pooled analysis of 31 studies of sleeve gastrectomy found anemia in 7% before the operation, 25% at three years and 18% at five years.[39]

The figures on deaths, repeat operations, readmissions and anemia after surgery are rated Emerging: they come from registries and pooled studies with differing definitions.

A GLP-1 drug after surgery

For people who lost less weight than expected after surgery, small randomized trials have tested adding a GLP-1 drug.[40][41] In one 68-week trial, 70 adults were randomly assigned and 63 were included in the analysis.[40] Those given semaglutide 2.4 mg lost about 18% of their weight and those given placebo gained 0.4%.[40]

The maker of semaglutide funded the 68-week trial and supplied the drug and placebo, with co-funding from a public research body and a charity.[40] A pooled analysis of 7 randomized trials of an older drug, liraglutide, in 460 people found more weight loss than placebo at 6 months and no statistically significant difference at 12 months.[41]

These results are rated Emerging: the trials were small, the semaglutide trial was funded by the drug's maker, and the pooled liraglutide result at 12 months was not statistically significant.

Wegovy injection (under the skin) and Wegovy tablets (by mouth) are FDA-approved for weight management in certain people; Ozempic injection, Ozempic tablets and Rybelsus tablets are approved only for adults with type 2 diabetes (checked September 27, 2026).

Wegovy injection and Zepbound injection each carry a boxed warning about thyroid C-cell tumors seen in rodents.

Sources: Wegovy injection prescribing information, Revised 06/2026; Zepbound injection prescribing information, Revised 08/2026.

What professional guidelines say

A 2022 joint statement from two bariatric surgery societies (ASMBS and IFSO) recommends metabolic and bariatric surgery for people with a BMI ≥35 kg/m2 'regardless of presence, absence, or severity of co-morbidities', and recommends considering it for people with metabolic disease and a BMI of 30 to 34.9 kg/m2.[42]

For Asian populations the statement recommends offering surgery at a BMI ≥27.5 kg/m2.[42] These are recommendations to clinicians.[42] Whether surgery is suitable for a person is a decision for that person and a surgical team.

A 2026 American College of Physicians living guideline suggests semaglutide or tirzepatide as first-line medicines, together with lifestyle changes, for nonpregnant adults with obesity, as a conditional recommendation based on moderate-certainty evidence for each.[43]

This page reports what studies found and is not advice about which treatment is right for you. A licensed clinician can review your health history and discuss both options with you.

FDA status of semaglutide and tirzepatide products

Approved uses differ by semaglutide product. Approved uses differ by tirzepatide product.

The FDA status table lists each semaglutide and tirzepatide product with its route and the uses in its prescribing information.

Ozempic injection, Ozempic tablets and Rybelsus tablets are not FDA-approved for weight management. Their approved uses are listed in the FDA status table. Prescribing them for weight management is off-label. For weight management in certain people, together with a reduced-calorie diet and more physical activity, FDA has approved Wegovy injection, given under the skin, and Wegovy tablets, taken by mouth.

Mounjaro injection is not FDA-approved for weight management. Its approved uses are listed in the FDA status table. Prescribing it for weight management is off-label. For weight management in certain adults, together with a reduced-calorie diet and more physical activity, FDA has approved Zepbound injection, given under the skin.

Each semaglutide product in the FDA status table is available only with a prescription (checked September 27, 2026).

Each tirzepatide product in the FDA status table is available only with a prescription (checked October 14, 2026).

Completing an intake form or making a payment does not guarantee a prescription.

Regulatory status checked September 27, 2026. Sources: www.accessdata.fda.gov/…/209637s038lbl.pdf; www.accessdata.fda.gov/…/213051Orig1s030lbl.pdf; www.accessdata.fda.gov/…/215256s025lbl.pdf; www.fda.gov/…/fda-clarifies-policies-compounders-national-glp-1-supply-begins-stabilize; public-inspection.federalregister.gov/…/2026-12937.pdf; www.fda.gov/…/fda-proposes-exclude-semaglutide-tirzepatide-and-liraglutide-503b-bulks-list; www.fda.gov/…/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss; www.fda.gov/…/compounding-and-fda-questions-and-answers; www.accessdata.fda.gov/…/218316Orig1s000ltr.pdf; www.accessdata.fda.gov/…/218316s007lbl.pdf

Regulatory status checked October 14, 2026. Sources: www.accessdata.fda.gov/…/215866s044s045lbl.pdf; www.accessdata.fda.gov/…/215866Orig1s044,215866Orig1s045ltr.pdf; www.accessdata.fda.gov/…/217806Orig1s041correctedlbl.pdf; www.fda.gov/…/fda-clarifies-policies-compounders-national-glp-1-supply-begins-stabilize; public-inspection.federalregister.gov/…/2026-08552.pdf; www.fda.gov/…/compounding-and-fda-questions-and-answers; www.fda.gov/…/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss

Provider directory

You can connect with a licensed provider to ask about an evaluation.

Find a Provider · Connect With a Licensed Provider

A licensed clinician decides whether treatment is appropriate. A prescription is not guaranteed.

Related pages

• Semaglutide for weight loss
• What is tirzepatide for weight loss?
• Semaglutide vs tirzepatide: what is the difference?
• How We Grade Medical Evidence