
==== Front
Br J Surg
Br J Surg
bjs
The British Journal of Surgery
0007-1323
1365-2168
Oxford University Press UK

10.1093/bjs/znae221
znae221
Original Article
AcademicSubjects/MED00910
Bjs/5
Cardiovascular and diabetes outcomes among patients with obesity and type 2 diabetes after metabolic bariatric surgery or glucagon-like peptide 1 receptor agonist treatment
https://orcid.org/0000-0001-9189-0093
Stenberg Erik Department of Surgery, Faculty of Medicine and Health, Örebro University, Örebro, Sweden

Ottosson Johan Department of Surgery, Faculty of Medicine and Health, Örebro University, Örebro, Sweden

Cao Yang Clinical Epidemiology and Biostatistics, School of Medical Sciences, Örebro University, Örebro, Sweden
Unit of Integrative Epidemiology, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden

Sundbom Magnus Department of Surgical Sciences, Uppsala University, Uppsala, Sweden

https://orcid.org/0000-0002-0166-6344
Näslund Erik Division of Surgery, Department of Clinical Sciences, Danderyd Hospital, Karolinska Institutet, Stockholm, Sweden

Correspondence to: Erik Stenberg, Department of Surgery, Faculty of Medicine and Health, Örebro University, Södra Grev Rosengatan, 70185 Örebro, Sweden (e-mail: erik.stenberg@regionorebrolan.se; @ErikStenberg_MD)
9 2024
05 9 2024
05 9 2024
111 9 znae22105 3 2024
05 6 2024
07 8 2024
© The Author(s) 2024. Published by Oxford University Press on behalf of BJS Foundation Ltd.
2024
https://creativecommons.org/licenses/by-nc/4.0/ This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com.

Abstract

Background

With the increasing prevalence of obesity and type 2 diabetes, the availability of different treatment options remains essential. Studies comparing the outcomes of glucagon-like peptide 1 receptor agonists with those of metabolic bariatric surgery in patients with type 2 diabetes and obesity are lacking.

Methods

Using propensity score matching, based on data from several nationwide clinical registries, patients who underwent primary metabolic bariatric surgery (Roux-en-Y gastric bypass or sleeve gastrectomy) were matched with patients who received glucagon-like peptide 1 receptor agonists. Outcome measures included the occurrence of major cardiovascular events, microvascular complications, and potential side effects (alcohol/substance abuse, self-harm, and fractures).

Results

Over a mean follow-up of 7 years, major cardiovascular events occurred in 191 of 2039 patients (cumulative incidence 14.5%) in the surgery group compared with 247 of 2039 patients (19.6%) in the glucagon-like peptide 1 receptor agonist group (HR 0.75 (95% c.i. 0.62 to 0.91), P = 0.003). Patients in the surgery group had lower haemoglobin A1c values 5 years after treatment (mean difference 9.82 (95% c.i. 8.51 to 11.14) mmol/mol, P < 0.001) and fewer microvascular complications (retinopathy HR 0.88 (95% c.i. 0.79 to 0.99), P = 0.039; nephropathy HR 0.72 (95% c.i. 0.66 to 0.80), P < 0.001; and neuropathy or leg ulcers HR 0.82 (95% c.i. 0.74 to 0.92), P < 0.001), but a higher risk of alcohol/substance abuse (HR 2.56 (95% c.i. 1.87 to 3.50), P < 0.001), self-harm (HR 1.41 (95% c.i. 1.17 to 1.71), P < 0.001), and fractures (HR 1.86 (95% c.i. 1.11 to 3.12), P = 0.019).

Conclusion

Compared with glucagon-like peptide 1 receptor agonist treatment, metabolic bariatric surgery is associated with superior metabolic outcomes and a lower risk of major cardiovascular events in patients with type 2 diabetes and obesity, but a higher risk of alcohol/substance abuse, self-harm, and fractures.

This propensity-score matched cohort study set out to compare the effects and risks of metabolic bariatric surgery with those of glucagon-like peptide 1 receptor agonist treatment in patients with type 2 diabetes and obesity. Over a mean follow-up of 7.1 years, surgery was associated with a reduced risk of major cardiovascular events and microvascular complications, but a higher risk of alcohol/substance abuse, self-harm, and fractures. Metabolic bariatric surgery should be considered in patients with severe obesity and type 2 diabetes, but the side effects, in the form of long-term risks of alcohol/substance abuse, self-harm, and fractures, require continued follow-up and close support.
==== Body
pmcIntroduction

In parallel with the current obesity pandemic, the prevalence of type 2 diabetes (T2D) has increased and this trend is estimated to continue in the near future1. Metabolic bariatric surgery (MBS) is associated with long-term weight loss, improvements in cardiometabolic co-morbidities, and improved life expectancy, particularly in patients with T2D or prediabetes2,3. Both observational studies and RCTs have shown that 60–70% of patients with obesity and T2D may be able to quit pharmacological treatment 5–7 years after MBS and 30–46% of patients have been reported to be in complete remission 5–10 years after surgery4–7. Similarly, MBS can result in hypertension remission8, with improvements in cardiovascular disease (CVD), including a reported 32% lower risk of major adverse cardiovascular events (MACE)9. Furthermore, MBS may play a role in secondary prevention of CVD, as suggested by a recent observational study10.

Recently, pharmacological treatment with glucagon-like peptide 1 receptor agonists (GLP-1RAs) has emerged as an alternative treatment for cardiometabolic diseases in patients with severe obesity and T2D. Liraglutide and dulaglutide have been found to reduce weight and lower the risk of MACE by 12–13% in patients with T2D and CVD, regardless of prior myocardial infarction11,12. Similarly, semaglutide has been found to lower the risk of MACE by 20–26% compared with placebo in patients with and without T2D13,14.

MBS is associated with improvements in T2D microvascular complications. The progression of retinopathy is halted in patients with mild disease and major renal outcomes are improved15. Similarly, GLP-1RA administration has been shown to prevent new-onset macroalbuminuria and glomerular filtration rate decline over time in patients with T2D16,17. Further, GLP-1RAs are associated with a lower risk of serious renal events compared with treatment with dipeptidyl peptidase 4 inhibitors18.

In a recent cohort study, MBS patients were associated with a lower risk of MACE compared with matched controls treated with GLP-1RAs19, but important clinical variables, such as BMI and T2D severity, were not taken into account. The aim of this study was to compare the effects of MBS and GLP-1RA therapy on macrovascular and microvascular disease, glycaemic and metabolic control, bodyweight, and potential side effects in a closely matched cohort of patients with obesity and T2D.

Methods

No pre-registration exists for this study.

This study was based on data from a cohort of patients, aged 18–70 years with T2D registered in the Scandinavian Obesity Surgery Registry (SOReg) or the Swedish National Diabetes Register (NDR) from 2007 to 2022. The SOReg is a national research and quality registry that started in 2007 and currently covers virtually all MBS procedures in Sweden. The registry has been continuously validated and is reported to have very high data validity20. The NDR started in 1996 and covers diagnostic and clinical information for nearly all individuals with diabetes in Sweden, with reports from specialized clinics and primary care facilities in Sweden21. By use of personal identification numbers (unique to all residents in Sweden), the registry was linked to: the Total Population Registry, covering all mortality and migration22; the National Patient Registry, a nationwide registry covering all hospital admissions since 1987, with an outpatient component that started in 2001, currently covering greater than 95% of outpatient visits in specialized healthcare23; the Prescribed Drugs Registry, established in 2005, including all dispensed, prescribed drugs classified according to the WHO’s Anatomic Therapeutic Chemical (ATC) classification system24; and the Swedish Cancer Registry, with greater than 96% completeness and very high morphological confirmation nationwide25. Socio-economic data were obtained from the Longitudinal Integrated Database for Health Insurance and Labour Market Studies26.

Definitions

T2D was defined as a previous diagnosis of T2D as registered in the NDR with ongoing pharmacological treatment for over 12 months. Cardiovascular co-morbidity was defined as a previous diagnosis of ischaemic heart disease, heart failure, or arrhythmic heart disease. Dyslipidaemia was defined as the use of lipid-modifying drugs. Chronic obstructive pulmonary disease (COPD) included moderate-to-severe COPD27. Smoking history was categorized as no smoking or a history of smoking. See Table S1.

Disposable income (total taxable income minus taxes and negative transfers) was adjusted for the 2023 consumer price index and divided into percentiles based on the disposable income of an age-, sex-, and county-matched cohort of the normal population.

The highest level of education was divided into three categories based on the highest level of education completed at the time of intervention: primary education, secondary education, or higher education.

Inclusion

Adults (aged 18–70 years) registered in the SOReg or the NDR with T2D and a BMI of 30–70  kg/m2, who either underwent a primary MBS procedure (Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG)) or received continuous GLP-1RA treatment (liraglutide, dulaglutide, or semaglutide) for a duration of greater than or equal to 3 years, were selected, as these interventions have been shown to reduce the risk of MACE in patients with T2D compared with placebo11–13. Missing information with regard to any of the inclusion criteria, only mono-pharmacological treatment for diabetes, and crossover between treatment arms were considered to be the exclusion criteria.

Matching

T2D patients who underwent a primary RYGB or SG procedure and who had not previously received GLP-1RA treatment were matched with T2D patients who had not been operated on and had received GLP-1RA treatment, where day of surgery or first day of GLP-1RA treatment was the index date. The matching was conducted using 1 : 1 propensity score matching with a generalized linear model and a caliper distance of 0.2, including age and BMI at the time of intervention, sex, cardiovascular co-morbidity, COPD, dyslipidaemia, treatment with antidepressant medication, history of cancer, haemoglobin A1c (HbA1c) value, duration of diabetes, number of diabetes drugs, insulin treatment, microvascular complications, disposable family income, smoking status, level of education, and year of intervention, stratified by baseline BMI (30–34.9 and greater than or equal to 35 kg/m2).

Outcomes

The main outcome for this study was the occurrence of MACE, defined as hospitalization for acute coronary syndrome, cerebrovascular events, or all-cause death. Secondary outcomes included bodyweight differences, remission of T2D (defined as HbA1c less than 48 mmol/mol without pharmacological treatment28), effect on insulin use, pharmacological treatment for dyslipidaemia reaching current treatment goals for dyslipidaemia29, new onset of other diabetes complications (nephropathy defined as estimated glomerular filtration rate less than 60 ml/min/1.73 m2 or albuminuria, retinopathy, or lower extremity complications defined as neuropathy or leg ulcers), occurrence of alcohol/substance abuse, self-harm, or fractures. See Table S2.

Statistics

Continuous variables are presented as mean(s.d.) or median (interquartile range (i.q.r.)), as appropriate. Categorical variables are presented as n (%). The risks of primary and secondary outcomes were evaluated using a multivariable Cox regression model, adjusted for all variables used for propensity score matching, with HRs and 95% confidence intervals as variance estimators accounting for clustering with matched pairs reported as measures of association. Patients were followed up until emigration, mortality, 10-year follow-up, or 31 December 2022, whichever came first. The time to event for longitudinal outcomes was estimated using the Kaplan–Meier method and is presented as cumulative probability (1 − Kaplan–Meier estimate). The effects on metabolic co-morbidities were evaluated using multivariable logistic regression (including all matching variables).

Differences in BMI loss, total weight loss (TWL), and HbA1c levels were analysed using treatment–effect estimators, which considered the propensity score matching. SPSS® (IBM, Armonk, NY, USA; version 29), Stata (StataCorp, College Station, TX, USA; version 17.0) and R (R Foundation for Statistical Computing, Vienna, Austria; version 4.2.0) were used for statistical analyses.

Results

During the study interval, 4499 patients who underwent surgery and 30 875 patients who were not operated on and were treated with GLP-1RAs for greater than or equal to 3 years were identified who matched the inclusion criteria.

Before matching, patients treated with GLP-1RAs were older, were more often men, and had more severe diabetes. See Table S3. The propensity score matching resulted in two similar groups with a standardized mean difference of less than 0.1 for all baseline characteristics. See Table 1 and Table S4. The mean(s.d.) follow-up time was 7.1(3.47) years in the surgery group and 7.0(3.10) years in the GLP-1RA group. The median duration of GLP-1RA treatment was 5.6 (i.q.r. 3.70–8.28) years.

Table 1 Baseline characteristics after propensity score matching

	Surgery group, n = 2039	GLP-1RA group, n = 2039	Standardized mean difference	
Age (years), mean(s.d.)	52.0(8.76)	51.8(10.9)	0.013	
BMI (kg/m2), mean(s.d.)	40.2(5.00)	40.2(5.61)	0.002	
Sex	
 Male	968 (47.5)	984 (48.3)	0.016	
 Female	1071 (52.5)	1055 (51.7)	0.016	
Co-morbidities	
 Cardiovascular	301 (14.8)	288 (14.1)	0.019	
 COPD	94 (4.6)	100 (4.9)	0.014	
 Dyslipidaemia	1242 (60.9)	1255 (61.5)	0.012	
 Depression	486 (23.8)	483 (23.7)	0.002	
Previous cancer	87 (4.3)	85 (4.2)	0.005	
Diabetes	
 HbA1c (mmol/mol), mean(s.d.)	63.1(16.99)	63.4(15.47)	0.017	
 Duration of diabetes (years), mean(s.d.)	7.7(5.96)	7.7(6.60)	0.011	
 Number of medications, mean(s.d.)	2.0(0.80)	2.1(0.66)	0.095	
 Insulin treatment	828 (40.6)	831 (40.8)	0.004	
 History of neuropathy or lower limb wounds	167 (8.2)	173 (8.5)	0.011	
 Retinopathy	521 (25.6)	494 (24.2)	0.032	
 Nephropathy	517 (25.4)	523 (25.6)	0.005	
Disposable income	
 Quartile 1	595 (29.2)	569 (27.9)	0.029	
 Quartile 2	623 (30.6)	655 (32.1)	0.032	
 Quartile 3	502 (24.6)	470 (23.1)	0.035	
 Quartile 4	319 (15.6)	345 (16.9)	0.035	
Education	
 Primary (9 years)	425 (20.8)	413 (20.3)	0.012	
 Secondary (10–12 years)	1164 (57.1)	1170 (57.4)	0.006	
 Higher (12 years)	450 (22.1)	456 (22.4)	0.007	
History of smoking	909 (44.6)	914 (44.8)	0.004	
Values are n (%) unless otherwise indicated. GLP-1RA, glucagon-like peptide 1 receptor agonist; COPD, chronic obstructive pulmonary disease; HbA1c, haemoglobin A1c.

The most common surgical procedure was RYGB (1669 patients; 82%), followed by SG (370 patients; 18%). The most commonly prescribed GLP-1RA at baseline was liraglutide (1589 patients; 78%), followed by semaglutide (273 patients; 13%) and dulaglutide (177 patients; 9%). Most patients received the same GLP-1RA throughout the study interval, although increased use of semaglutide was seen over time (at 3 years: 509 patients; 25%).

Cardiovascular endpoints

Over a 10-year follow-up interval, 191 patients experienced MACE in the surgery group (cumulative incidence 14.5% (95% c.i. 12.46% to 16.63%)) compared with 247 patients in the GLP-1RA group (cumulative incidence 19.6% (95% c.i. 17.13% to 21.98%)) (HR 0.75 (95% c.i. 0.62 to 0.91), P = 0.003). See Fig. 1 and Table 2. Non-fatal acute cardiac events occurred less frequently in the surgery group (51 ???, cumulative incidence 4.2% (95% c.i. 2.70% to 5.42%)) compared with the GLP-1RA group (112 ???, cumulative incidence 8.6% (95% c.i. 6.94% to 10.33%)) (HR 0.46 (95% c.i. 0.33 to 0.65), P < 0.001), whereas no difference was seen for non-fatal cerebrovascular events (40 events, cumulative incidence 2.8% (95% c.i. 1.89% to 3.78%) in the surgery group versus 46 events, cumulative incidence 4.2% (95% c.i. 2.81% to 5.49%) the GLP-1RA group) (HR 0.87 (95% c.i. 0.56 to 1.34), P = 0.521).

Fig. 1 Ten-year cumulative probabilities for major cardiovascular events, mortality, new-onset nephropathy, and new-onset retinopathy (comparing patients with type 2 diabetes who underwent surgery with those treated with glucagon-like peptide 1 receptor agonists)

a Ten-year cumulative probability for major cardiovascular events. b Ten-year cumulative probability for mortality. c Ten-year cumulative probability for new-onset nephropathy. d Ten-year cumulative probability for new-onset retinopathy. GLP-1RA, glucagon-like peptide 1 receptor agonist.

Table 2 Major and secondary outcomes for metabolic bariatric surgery compared with GLP-1RA treatment

	Surgery group, n (cumulative incidence; 95% c.i.)	GLP-1RA group, n (cumulative incidence; 95% c.i.)	Absolute risk difference (95% c.i.)	HR (95% c.i.)	P	
MACE	191 (14.5; 12.5,16.6)	247 (19.6%; 17.1%,22.0%)	−5.1 (−8.0,−2.2)	0.75 (0.62,0.91)	0.003	
Non-fatal acute cardiac events	51 (4.2; 2.7,5.4)	112 (8.6; 6.9,10.3)	−4.4 (−6.3,−2.5)	0.46 (0.33,0.65)	<0.001	
Non-fatal cerebrovascular events	40 (2.8; 1.9,3.8)	46 (4.2; 2.8,5.5)	−1.4 (−2.8,0.1)	0.87 (0.56,1.34)	0.521	
Ten-year mortality	120 (9.6; 7.9,11.4)	123 (10.7; 8.8,12.7)	−1.1 (−3.5,1.3)	0.97 (0.75,1.24)	0.801	
Ten-year mortality (excluding deaths within 3 years)	99 (8.7; 6.9,10.3)	121 (10.7; 8.7,12.6)	−2.0 (−4.4,0.4)	0.79 (0.61,1.03)	0.078	
Microvascular complications	
 Retinopathy*	265 (39.3; 34.8,43.8)	409 (52.8; 48.1,57.4)	−13.4 (−18.6,−8.4)	0.88 (0.79,0.99)	0.039	
 Nephropathy*	350 (44.3; 39.8,48.8)	529 (58.4; 54.0,62.8)	−14.1 (−18.8,−9.4)	0.72 (0.66,0.80)	<0.001	
 Estimated glomerular filtration rate <60 ml/min/1.73 m2	271 (25.1; 21.8,28.4)	446 (33.4; 30.2,36.7)	−8.3 (−11.9,−4.7)	0.65 (0.56,0.76)	<0.001	
 Neuropathy or leg ulcers*	391 (39.9; 36.1,43.7)	538 (47.5; 43.6,51.4)	−7.6 (−11.8,−3.4)	0.82 (0.74,0.92)	<0.001	
Other complications	
 Alcohol/substance abuse	139 (9.7; 8.1,11.4)	57 (3.7; 2.7,4.7)	6.0 (4.2,7.8)	2.56 (1.87,3.50)	<0.001	
 Self-harm	42 (2.9; 2.0,3.8)	24 (1.7; 1.0,2.5)	1.2 (0.1,2.3)	1.41 (1.17,1.71)	<0.001	
 Fractures	260 (20.6; 18.1,23.0)	184 (14.0; 11.9,16.1)	6.6 (6.6,9.5)	1.86 (1.11,3.12)	0.019	
Cumulative incidence and absolute risk difference values are given as percentages. *Analysis only includes patients without this complication at baseline. Excluded from analysis: retinopathy, 521 patients in the surgery group and 494 patients in the glucagon-like peptide 1 receptor agonist group; nephropathy, 521 patients in the surgery group and 494 patients in the glucagon-like peptide 1 receptor agonist group; and neuropathy or leg ulcers, 517 patients in the surgery group and 523 patients in the glucagon-like peptide 1 receptor agonist group. GLP-1RA, glucagon-like peptide 1 receptor agonist; MACE, major cardiovascular events.

Mortality

In the surgery group, one patient died from an infectious complication within 90 days of surgery (90-day mortality rate 0.05%). During the 10 years after intervention, 120 patients died in the surgery group (cumulative incidence 9.6% (95% c.i. 7.87% to 11.41%)) compared with 123 patients in the GLP-1RA group (cumulative incidence 10.7% (95% c.i. 8.81% to 12.68%)) (HR 0.97 (95% c.i. 0.75 to 1.24), P = 0.801). The main causes of death are presented in Table 3. After exclusion of cases and controls with death within 3 years after intervention there were 99 deaths in the surgery group (cumulative incidence 8.7% (95% c.i. 6.92% to 10.29%)) and 121 deaths in the GLP-1RA group (cumulative incidence 10.7% (95% c.i. 8.73% to 12.60%)) (HR 0.79 (95% c.i. 0.61 to 1.03), P = 0.078).

Table 3 Mortality up to 10 years after intervention, excluding cases and controls in the event of a death within 3 years after intervention

Mortality	Surgery group, n = 99	GLP-1RA group, n = 121	
Cardiovascular	35	52	
Cancer	25	31	
Infection	8	11	
Suicide, intoxication, or accidents	13	7	
Respiratory disorder	2	2	
Liver disease	7	7	
Kidney disease	2	5	
Other causes	7	6	
Values are n. GLP-1RA, glucagon-like peptide 1 receptor agonist.

Weight

When compared with the BMI before intervention (start of GLP-1RA treatment or preoperative weight reduction before MBS), the mean(s.d.) BMI loss at 2 years was 10.5(4.19) kg/m2 in the surgery group compared with 1.7(2.96) kg/m2 in the GLP-1RA group (mean difference 8.75 (95% c.i. 8.47 to 9.04) kg/m2, P < 0.001); the corresponding mean(s.d.) TWL was 25.9(8.98)% and 4.3(6.49)% respectively (mean difference 21.67% (95% c.i. 21.05% to 22.29%), P < 0.001). The mean(s.d.) BMI loss at 5 years was 9.4(4.24) kg/m2 in the surgery group compared with 2.1(3.48) kg/m2 in the GLP-1RA group (mean difference 6.97 (95% c.i. 6.62 to 7.32) kg/m2, P < 0.001); the corresponding mean(s.d.) TWL was 23.3(9.39)% and 5.3(7.80)% respectively (mean difference 17.27% (95% c.i. 16.46% to 18.07%), P < 0.001).

See Fig. 2a.

Fig. 2 Five-year trajectories for weight and haemoglobin A1c (comparing patients with type 2 diabetes who underwent surgery with those treated with glucagon-like peptide 1 receptor agonists)

a Five-year trajectory for weight (kg). b Five-year trajectory for haemoglobin A1c (mmol/mol). GLP-1RA, glucagon-like peptide 1 receptor agonist.

Glycaemic control

At 2 years, the mean(s.d.) HbA1c was 44.9(11.81) mmol/mol in the surgery group and 56.6(15.91) mmol/mol in the GLP-1RA group (mean difference 12.11 (95% c.i. 11.05 to 13.17) mmol/mol, P < 0.001). Controlled diabetes with HbA1c less than 53 mmol/mol was reported for 1217 of 1497 patients (81.3%) in the surgery group compared with 837 of 1734 patients (48.3%) in the GLP-1RA group (P < 0.001); 209 patients (11.1%) in the surgery group still received insulin compared with 794 patients (39.0%) in the GLP-1RA group (P < 0.001).

At 5 years, the mean(s.d.) HbA1c was 50.2(13.14) mmol/mol in the surgery group and 60.4(15.86) mmol/mol in the GLP-1RA group (mean difference 9.82 (95% c.i. 8.51 to 11.14) mmol/mol, P < 0.001). Controlled diabetes with HbA1c less than 53 mmol/mol was reported for 689 of 1034 patients (66.6%) in the surgery group compared with 406 of 1147 patients (35.4%) in the GLP-1RA group (P < 0.001). Complete remission of T2D was reported for 331 of 1034 patients (32.0%) in the surgery group compared with 4 of 1147 patients (0.3%) in the GLP-1RA group (P < 0.001); 172 patients (12.4%) in the surgery group still received insulin compared with 737 patients (54.0%) in the GLP-1RA group (P < 0.001).

See Fig. 2b.

Dyslipidaemia

At 2 years, the use of medication for dyslipidaemia was lower in the surgery group (40.1%) compared with the GLP-1RA group (68.1%) (P < 0.001) and 265 patients in the surgery group (17.6%) reached the goals for lipid status as set by the American Diabetes Association with or without medication compared with 145 patients in the GLP-1RA group (8.4%) (P < 0.001).

At 5 years, 44.4% of the patients in the surgery group used medication for dyslipidaemia compared with 72.4% of the patients in the GLP-1RA group (P < 0.001). In the surgery group, 203 patients (19.0%) reached the goals for lipid status as set by the American Diabetes Association with or without medication compared with 101 patients (8.6%) in the GLP-1RA group (P < 0.001).

Other endpoints

The surgery group had significantly fewer microvascular complications (such as retinopathy, nephropathy, and neuropathy) compared with the GLP-1RA group at follow-up. In contrast, patients in the GLP-1RA group were less likely to be diagnosed with alcohol/substance abuse, self-harm, and fractures compared with patients in the surgery group. See Table 2.

Subgroup analyses

Similar effects on diabetes remission, mean HbA1c, and microvascular complications were seen both in patients with a BMI of 30–34.9 kg/m2 and in patients with a BMI of greater than or equal to 35 kg/m2. Although the group with a BMI of 30–34.9 kg/m2 was too small to estimate effects on mortality, MACE, alcohol/substance abuse, and self-harm, similar trends were seen as in patients with a BMI greater than or equal to 35 kg/m2 for whom surgery was associated with a reduced risk of MACE, but a higher risk of alcohol/substance abuse and self-harm. See Tables S5, S6.

Discussion

In this propensity-score matched cohort study of patients with obesity and T2D, MBS is associated with superior outcomes regarding weight loss, control and remission of T2D, reduction in MACE, and microvascular complications compared with GLP-1RA treatment used at a dosage for diabetes control. The other side of the coin for this metabolic improvement is an increased risk of alcohol/substance abuse, self-harm, and fractures over a 10-year follow-up interval.

The 5-year weight loss after MBS observed in this study is similar to that reported previously30. Weight loss is often lower in patients with T2D30. The two most used GLP-1RA medications in this study are liraglutide and semaglutide; semaglutide has been shown to be superior to liraglutide in terms of weight loss31. In a large cohort study of patients with obesity and T2D, a 20% weight loss was required in the non-surgical group to achieve a reduced risk of MACE, whereas a 10% weight loss was required in the surgical group32, and thus the weight loss seen after GLP-1RA treatment may not reach the weight loss threshold required to achieve a reduced risk of MACE.

MBS has consistently been shown to have a superior effect compared with drug therapy in controlling T2D33. The rate of complete remission for T2D observed in this study is lower than that reported in many other observational studies and this may be attributed to varying definitions of T2D used by different studies. The definition used in the present study excludes milder forms of T2D. Many observational studies rely solely on a reduction in the use of anti-diabetic medications to define T2D remission, whereas, in the present study, T2D remission is defined as both the cessation of anti-diabetic medications and a normalization of HbA1c levels.

MBS has previously been shown to reduce MACE and decrease mortality compared with non-surgical care in patients with T2D9. However, in the study by Aminian et al.9, it is not clear how many patients in the non-surgically treated group were treated with GLP-1RAs. The cumulative incidence rates of MACE at 10-year follow-up in the present study were slightly lower than those reported by Aminian et al.9 at 8-year follow-up in both the surgical and non-surgical groups. This may be attributed to differences in the types of surgical procedures employed, the use of GLP-1RAs, baseline BMI values, and co-morbidities.

A meta-analysis of several different GLP-1RAs found that GLP-1RA treatment in patients with T2D reduced MACE and all-cause mortality compared with placebo34,35. This was true for liraglutide11, dulaglutide12, and semaglutide13. Overall, significant reductions in the range of 9–16% in the incidence of myocardial infarction, stroke, cardiovascular and all-cause death have been seen after treatment36. Semaglutide has also been shown to reduce the risk of death due to cardiovascular causes in patients with pre-existing CVD and obesity, but without T2D14. GLP-1RAs modify several risk factors for cardiovascular complications, including inducing weight loss, lowering blood pressure, reducing plasma low density lipoprotein cholesterol and triglyceride concentrations, and improving glycaemic control, which may contribute to the observed reduction in MACE37. MBS also affects these factors, but to a greater degree, as shown by greater weight loss, better glycaemic control, and a higher proportion of patients reaching treatment goals for serum lipids in this study. This is coupled with the fact that MBS also increases postprandial GLP-1 plasma concentrations38, which may account for its greater impact on MACE.

MBS is also associated with fewer microvascular complications than GLP-1RA treatment. GLP-1RA treatment has been shown to prevent new-onset macroalbuminuria and slow the reduction in glomerular filtration rate36. A recent meta-analysis of microvascular outcomes after MBS in patients with obesity and T2D demonstrated that end-stage renal disease was reduced by 69% and neuropathy by 59%39 and this is in line with the results of the present study. This difference in the outcomes of microvascular disease post-MBS may also be attributed to the greater impact on weight loss and glycaemic control compared with GLP-1RA treatment. However, surgery is associated with a greater risk of alcohol/substance abuse, self-harm, and fractures compared with GLP-1RA treatment.

This study has several limitations. First, despite using a comprehensive matching process, residual measured or unmeasured confounders, such as the choice of treatment modality exercised by the patient, may have influenced the findings. Data on compliance with recommended supplementation, which may influence the risk of fractures, was not available. Second, coding errors, misclassifications, and misdiagnoses in the registry data may have influenced the outcomes. However, the included registries have been continuously validated with very high data validity. Third, most patients were treated with older versions of GLP-1RAs. Newer GLP-1RAs, such as semaglutide, have been found to have a greater impact on weight loss31. However, an analysis of only newer GLP-1RAs would be limited by a shorter follow-up interval and this may prevent the detection of long-term changes. Finally, patients in the GLP-1RA group completed at least 3 years of treatment. Hence, patients with low compliance rates and those experiencing significant side effects were not included in the study. In addition, this requirement for inclusion in the GLP-1RA group meant that anyone who died within 3 years after treatment was not considered for inclusion in the study, whereas such patients in the surgery group were available for inclusion. Although the mortality rate remained very low during the first few years after surgery, this requirement will likely overestimate mortality rates in the surgical group. Given the hypothesis that surgery is associated with a reduced occurrence of MACE, the authors accepted this bias that is likely to underestimate the treatment effects of MBS. The major strengths of this study are the high quality and completeness of the data and careful matching of the available covariates.

Compared with GLP-1RA treatment, MBS is associated with superior metabolic outcomes and a lowered risk of MACE in patients with T2D and obesity. However, MBS is associated with a higher risk of alcohol/substance abuse, self-harm, and fractures.

Supplementary Material

znae221_Supplementary_Data

Funding

This work was supported by grants from Region Örebro County (OLL-939106), Region Stockholm, and the Strategic Research Programme in Diabetes. The funders had no role in: the design and conduct of the study; the collection, management, analysis, and interpretation of the data; the preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.

Disclosure

E.S. has received reimbursement for lectures from Johnson & Johnson Medical and MSD, and consultant fees from Johnson & Johnson Medical (paid to institution). J.O. is a member of the safety committee for the BEST study. M.S. is part of the project group for the National Guidelines for Obesity Care run by the National Board of Health and Welfare. The authors declare no other conflict of interest.

Supplementary material

Supplementary material is available at BJS online.

Data availability

Data cannot be shared publicly because of patient confidentiality under current Swedish legislation. Data are available from the Scandinavian Obesity Surgery Registry (contact via soreg@regionorebrolan.se), the Swedish Board of Health and Welfare (contact via Registerservice@socialstyrelsen.se), and Statistics Sweden (contact via mikrodata@scb.se) for researchers who meet the criteria for access to confidential data.

Role of the funder

The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Author contributions

Erik Stenberg (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Visualization, Writing—original draft), Johan Ottosson (Conceptualization, Data curation, Validation, Writing—review & editing), Yang Cao (Data curation, Formal analysis, Investigation, Methodology, Software, Writing—review & editing), Magnus Sundbom (Conceptualization, Project administration, Validation, Writing—review & editing), and Erik Näslund (Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing—review & editing)
==== Refs
References

1 Andersson  T, Ahlbom  A, Carlsson  S. Diabetes prevalence in Sweden at present and projections for year 2050. PLoS One  2015;10 :e0143084 26619340
2 Syn  NL, Cummings  DE, Wang  LZ, Lin  DJ, Zhao  JJ, Loh  M  et al  Association of metabolic-bariatric surgery with long-term survival in adults with and without diabetes: a one-stage meta-analysis of matched cohort and prospective controlled studies with 174 772 participants. Lancet  2021;397 :1830–1841 33965067
3 Arterburn  DE, Telem  DA, Kushner  RF, Courcoulas  AP. Benefits and risks of bariatric surgery in adults: a review. JAMA  2020;324 :879–887 32870301
4 Mingrone  G, Panunzi  S, De Gaetano  A, Guidone  C, Iaconelli  A, Capristo  E  et al  Metabolic surgery versus conventional medical therapy in patients with type 2 diabetes: 10-year follow-up of an open-label, single-centre, randomised controlled trial. Lancet  2021;397 :293–304 33485454
5 Schauer  PR, Bhatt  DL, Kirwan  JP, Wolski  K, Aminian  A, Brethauer  SA  et al  Bariatric surgery versus intensive medical therapy for diabetes—5-year outcomes. N Engl J Med  2017;376 :641–651 28199805
6 Backman  O, Bruze  G, Näslund  I, Ottosson  J, Marsk  R, Neovius  M  et al  Gastric bypass surgery reduces de novo cases of type 2 diabetes to population levels: a nationwide cohort study from Sweden. Ann Surg  2019;269 :895–902 30102631
7 Jans  A, Näslund  I, Ottosson  J, Szabo  E, Näslund  E, Stenberg  E. Duration of type 2 diabetes and remission rates after bariatric surgery in Sweden 2007–2015: a registry-based cohort study. PLoS Med  2019;16 :e1002985 31747392
8 Schiavon  CA, Bersch-Ferreira  AC, Santucci  EV, Oliveira  JD, Torreglosa  CR, Bueno  PT  et al  Effects of bariatric surgery in obese patients with hypertension: the GATEWAY randomized trial (gastric bypass to treat obese patients with steady hypertension). Circulation  2018;137 :1132–1142 29133606
9 Aminian  A, Zajichek  A, Arterburn  DE, Wolski  KE, Brethauer  SA, Schauer  PR  et al  Association of metabolic surgery with major adverse cardiovascular outcomes in patients with type 2 diabetes and obesity. JAMA  2019;322 :1271–1282 31475297
10 Näslund  E, Stenberg  E, Hofmann  R, Ottosson  J, Sundbom  M, Marsk  R  et al  Association of metabolic surgery with major adverse cardiovascular outcomes in patients with previous myocardial infarction and severe obesity: a nationwide cohort study. Circulation  2021;143 :1458–1467 33103469
11 Marso  SP, Daniels  GH, Brown-Frandsen  K, Kristensen  P, Mann  JF, Nauck  MA  et al  Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med  2016;375 :311–322 27295427
12 Gerstein  HC, Colhoun  HM, Dagenais  GR, Diaz  R, Lakshmanan  M, Pais  P  et al  Dulaglutide and cardiovascular outcomes in type 2 diabetes (REWIND): a double-blind, randomised placebo-controlled trial. Lancet  2019;394 :121–130 31189511
13 Marso  SP, Bain  SC, Consoli  A, Eliaschewitz  FG, Jódar  E, Leiter  LA  et al  Semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med  2016;375 :1834–1844 27633186
14 Lincoff  AM, Brown-Frandsen  K, Colhoun  HM, Deanfield  J, Emerson  SS, Esbjerg  S  et al  Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med  2023;389 :2221–2232 37952131
15 Bashir  B, Iqbal  Z, Adam  S, Ferdousi  M, Chick  W, Hussein  HA  et al  Microvascular complications of obesity and diabetes—role of bariatric surgery. Obes Rev  2023;24 :e13602 37515402
16 Mann  JFE, Ørsted  DD, Brown-Frandsen  K, Marso  SP, Poulter  NR, Rasmussen  S  et al  Liraglutide and renal outcomes in type 2 diabetes. N Engl J Med  2017;377 :839–848 28854085
17 Heerspink  HJL, Apperloo  E, Davies  M, Dicker  D, Kandler  K, Rosenstock  J  et al  Effects of semaglutide on albuminuria and kidney function in people with overweight or obesity with or without type 2 diabetes: exploratory analysis from the STEP 1, 2, and 3 trials. Diabetes Care  2023;46 :801–810 36801984
18 Pasternak  B, Wintzell  V, Eliasson  B, Svensson  AM, Franzen  S, Gudbjornsdottir  S  et al  Use of glucagon-like peptide 1 receptor agonists and risk of serious renal events: Scandinavian cohort study. Diabetes Care  2020;43 :1326–1335 32295809
19 Stenberg  E, Näslund  E. Major adverse cardiovascular events among patients with type-2 diabetes, a nationwide cohort study comparing primary metabolic and bariatric surgery to GLP-1 receptor agonist treatment. Int J Obes (Lond)  2023;47 :251–256 36670155
20 Sundbom  M, Näslund  I, Näslund  E, Ottosson  J. High acquisition rate and internal validity in the Scandinavian Obesity Surgery Registry. Surg Obes Relat Dis  2021;17 :606–614 33243667
21 Eliasson  B, Gudbjörnsdottir  S. Diabetes care – improvement through measurement. Diabetes Res Clin Pract  2014; 106 (Suppl 2 ): S291–S294 25550056
22 Ludvigsson  JF, Almqvist  C, Bonamy  AK, Ljung  R, Michaëlsson  K, Neovius  M  et al  Registers of the Swedish total population and their use in medical research. Eur J Epidemiol  2016;31 :125–136 26769609
23 Ludvigsson  JF, Andersson  E, Ekbom  A, Feychting  M, Kim  JL, Reuterwall  C  et al  External review and validation of the Swedish National Inpatient register. BMC Public Health  2011;11 :450 21658213
24 Wallerstedt  SM, Wettermark  B, Hoffmann  M. The first decade with the Swedish Prescribed Drug Register – a systematic review of the output in the scientific literature. Basic Clin Pharmacol Toxicol  2016;119 :464–469 27112967
25 Barlow  L, Westergren  K, Holmberg  L, Talbäck  M. The completeness of the Swedish Cancer Register: a sample survey for year 1998. Acta Oncol  2009;48 :27–33 18767000
26 Ludvigsson  JF, Svedberg  P, Olén  O, Bruze  G, Neovius  M. The Longitudinal Integrated Database for Health Insurance and Labour Market Studies (LISA) and its use in medical research. Eur J Epidemiol  2019;34 :423–437 30929112
27 Riley  CM, Sciurba  FC. Diagnosis and outpatient management of chronic obstructive pulmonary disease: a review. JAMA  2019;321 :786–797 30806700
28 Riddle  MC, Cefalu  WT, Evans  PH, Gerstein  HC, Nauck  MA, Oh  WK  et al  Consensus report: definition and interpretation of remission in type 2 diabetes. Diabetes Care  2021;44 :2438–2444 34462270
29 American Diabetes Association . 10. Cardiovascular disease and risk management: Standards of Medical Care in Diabetes—2021. Diabetes Care  2021; 44 (Suppl 1 ): S125–S150 33298421
30 Arterburn  D, Wellman  R, Emiliano  A, Smith  SR, Odegaard  AO, Murali  S  et al  Comparative effectiveness and safety of bariatric procedures for weight loss: a PCORnet cohort study. Ann Intern Med  2018;169 :741–750 30383139
31 Stretton  B, Kovoor  J, Bacchi  S, Chang  S, Ngoi  B, Murray  T  et al  Weight loss with subcutaneous semaglutide versus other glucagon-like peptide 1 receptor agonists in type 2 diabetes: a systematic review. Intern Med J  2023;53 :1311–1320 37189293
32 Aminian  A, Zajichek  A, Tu  C, Wolski  KE, Brethauer  SA, Schauer  PR  et al  How much weight loss is required for cardiovascular benefits? Insights from a metabolic surgery matched-cohort study. Ann Surg  2020;272 :639–645 32932320
33 Kirwan  JP, Courcoulas  AP, Cummings  DE, Goldfine  AB, Kashyap  SR, Simonson  DC  et al  Diabetes remission in the Alliance of Randomized Trials of Medicine versus Metabolic Surgery in Type 2 Diabetes (ARMMS-T2D). Diabetes Care  2022;45 :1574–1583 35320365
34 Kristensen  SL, Rorth  R, Jhund  PS, Docherty  KF, Sattar  N, Preiss  D  et al  Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of cardiovascular outcome trials. Lancet Diabetes Endocrinol  2019;7 :776–785 31422062
35 Husain  M, Bain  SC, Jeppesen  OK, Lingvay  I, Sorrig  R, Treppendahl  MB  et al  Semaglutide (SUSTAIN and PIONEER) reduces cardiovascular events in type 2 diabetes across varying cardiovascular risk. Diabetes Obes Metab  2020;22 :442–451 31903692
36 Nauck  MA, Quast  DR, Wefers  J, Meier  JJ. GLP-1 receptor agonists in the treatment of type 2 diabetes – state-of-the-art. Mol Metab  2021;46 :101102 33068776
37 Nauck  MA, Meier  JJ, Cavender  MA, Abd El Aziz  M, Drucker  DJ. Cardiovascular actions and clinical outcomes with glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors. Circulation  2017;136 :849–870 28847797
38 Falken  Y, Hellstrom  PM, Holst  JJ, Naslund  E. Changes in glucose homeostasis after Roux-en-Y gastric bypass surgery for obesity at day three, two months, and one year after surgery: role of gut peptides. J Clin Endocrinol Metab  2011;96 :2227–2235 21543426
39 Obeso-Fernández  J, Millan-Alanis  JM, Rodríguez-Bautista  M, Medrano-Juarez  S, Oyervides-Fuentes  S, Gonzalez-Cruz  D  et al  Benefits of bariatric surgery on microvascular outcomes in adult patients with type 2 diabetes: a systematic review and meta-analysis. Surg Obes Relat Dis  2023;19 :916–927 37169666
