Abstract
OBJECTIVE
To assess the effect of metabolic bariatric surgery (MBS) in people living with type 1 diabetes (T1D) and obesity.
RESEARCH DESIGN AND METHODS
From retrospective multicenter data, total weight loss, daily insulin requirement, HbA1c, and cardiometabolic parameters were assessed before and after surgery. Longitudinal models were used to identify response determinants.
RESULTS
This study included 162 people living with T1D and obesity. One year after surgery, mean total weight loss percentage was 29.7% (interquartile range [IQR] = 29.4, 30.3). Insulin requirements dropped from 0.75 (IQR = 0.58, 1.00) units/kg/day to 0.32 (IQR = 0.23, 0.43) units/kg/day (P < 0.001), and HbA1c dropped from 64.0 (IQR = 57.0, 74.0) mmol/mol to 60.0 (IQR = 53.4, 68.0) mmol/mol (P < 0.001). LDL, HDL, total cholesterol, and triglyceride levels significantly improved after surgery (P < 0.001). Greater total weight loss was associated with reduced insulin requirements, and higher baseline HbA1c was associated with poorer postsurgery glycemic control.
CONCLUSIONS
MBS was associated with substantial metabolic improvement in people with T1D and obesity, in particular in those with high HbA1c and insulin need at baseline.
Graphical Abstract
Introduction
Overweight and obesity affect up to 60% of people with type 1 diabetes (T1D) and complicate glycemic management by increasing insulin requirements (1–3). Metabolic bariatric surgery (MBS) reduces body weight and insulin needs, but the effect on glycemic control remains a matter of debate (4–8). In this multicenter retrospective study, we evaluated changes in body weight, insulin requirements, HbA1c, lipid profile (LDL, HDL, total cholesterol, and triglycerides), blood pressure, and kidney function during the first year after surgery in adults with T1D and obesity.
Research Design and Methods
Study Design and Population
Retrospective data were obtained from electronic clinical records from four centers in the Netherlands, three in Belgium, and one in Kuwait. Adults (aged >18 years) with T1D (confirmed by C-peptide, GAD and/or IA2 levels, or clinical diagnosis) and obesity (BMI >30) were included. All procedures followed national eligibility criteria and local ethical standards (details are provided in Supplementary Methods).
Outcomes
Body weight, BMI, and percentage total weight loss (TWL) were assessed preoperatively and up to 15 months after surgery. Glycemic outcomes included HbA1c (mmol/mol) and total daily insulin dose (units/day, units/kg/day). Information of insulin pump type, including automated insulin delivery (AID) versus nonautomated pump therapy, was extracted where available. Additional metabolic parameters included lipid profile (LDL, HDL, total cholesterol, and triglyceride levels), renal function (according to the Chronic Kidney Disease Epidemiology Collaboration equation), and blood pressure.
Information on peri- and postoperative complications within the first 30 days, including bleeding, gastrointestinal leak, and diabetic ketoacidosis, was retrieved from medical records. Due to nonstandardized data collection and intercenter variability, statistical analyses were not performed for these outcomes.
Statistical Analysis
Statistical analyses were conducted using R, version 4.2.1. Descriptive statistics summarized baseline characteristics and postoperative changes; continuous data are presented as mean ± SD or median (IQR), depending on distribution, and categorical variables are presented as counts and percentages. Paired t tests or Wilcoxon signed-rank tests compared pre- and postoperative values, with Holm-Bonferroni correction for multiple testing. Missingness was handled with multiple imputation with chained equations (M = 100). Analyses were performed in each imputed data set and pooled using Rubin’s rules.
Longitudinal changes in TWL, HbA1c, and daily insulin requirements were analyzed using generalized additive models (GAMs) and linear mixed models (LMMs). Only Dutch and Belgian data were included, due to limited Kuwaiti follow-up. Each model included outcome-specific covariates (e.g., sex, procedure type, insulin delivery method, baseline HbA1c, TWL). Model selection was based on fit statistics (e.g., R2) and included both fixed effects and smooth or random terms (e.g., time, patient identifier), as appropriate. Sensitivity analyses excluding clinically diagnosed T1D were performed.
Results
Clinical Characteristics
A total of 177 individuals with T1D and obesity were screened, of whom 15 were excluded (detailed information is provided in Supplementary Methods). The final cohort consisted of 162 individuals, of whom 77.8% were women. Among cohort members, 42.0% underwent Roux-en-Y gastric bypass (RYGB) and 58.0% underwent sleeve gastrectomy (SG). The majority used multiple daily injections (MDI) for insulin delivery (57.4%). Baseline and 12-month clinical and metabolic characteristics are summarized in Table 1, center-specific data are listed in Supplementary Table 1 and absolute and relative changes are presented in Table 2.
Table 1.
Baseline characteristics and postoperative outcomes
| Characteristic | Baseline | 12 months after surgery | P value* |
|---|---|---|---|
| No. of participants | 162 | ||
| Female sex, n (%) | 126 (77.8) | ||
| Age at surgery (years) | 37.5 [29.1, 49.7] | ||
| Race and ethnicity, n (%) | |||
| White/European | 124 (77.0) | ||
| Arab | 38 (33.0) | ||
| SG procedure, n (%) | 94 (58.0) | ||
| Diabetes duration (years) | 20.0 [11.0, 27.3] | ||
| MDI insulin delivery, n (%) | 93 (57.4) | ||
| Type of insulin pump, n (%) | |||
| Non-AID | 47 (69.1) | ||
| AID | 15 (22.1) | ||
| Unknown | 6 (8.8) | ||
| Weight (kg) | 113.6 (20.9) | 80.3 (16.6) | <0.001 |
| BMI (kg/m²) | 40.2 (5.6) | 28.5 (4.8) | <0.001 |
| Waist circumference (cm) | 120.6 (16.3) | 99.5 (13.0) | <0.001 |
| HbA1c (mmol/mol) | 64.0 [57.0, 74.0] | 60.0 [53.4, 68.0] | <0.001 |
| Insulin total | |||
| Units/day | 82.6 [60.0, 115.3] | 36.3 [24.0, 50.1] | <0.001 |
| Units/kg/day | 0.75 [0.58, 1.00] | 0.32 [0.23, 0.43] | <0.001 |
| Hemoglobin (mmol/L) | 8.5 (0.9) | 8.1 (1.0) | <0.001 |
| eGFR (mL/min/1.73 m²) | 96.5 (23.7) | 94.3 (21.3) | 0168 |
| Cholesterol (mmol/L) | 4.84 (1.00) | 4.21 (0.87) | <0.001 |
| LDL (mmol/L) | 2.63 [2.20, 3.41] | 2.24 [1.80, 2.60] | <0.001 |
| HDL (mmol/L) | 1.21 [1.01, 1.48] | 1.44 [1.16, 1.74] | <0.001 |
| Triglycerides (mmol/L) | 1.27 [0.94, 2.10] | 0.80 [0.67, 1.02] | <0.001 |
| SBP (mmHg) | 136.5 (18.5) | 125.0 (19.0) | <0.001 |
| DBP (mmHg) | 81.5 (11.2) | 75.5 (10.4) | <0.001 |
| Uses hypertension medication | 45 (27.8) | 28 (17.3); M = 11 | 0.039 |
| Uses lipid-lowering medication | 45 (27.8); M = 10 | 52 (32.1); M = 19 | 0.045 |
Values are displayed as mean (SD) for normally distributed data and median [IQR] for nonnormally distributed data, unless otherwise indicated. Statistical tests were paired t test and Wilcoxon rank or McNemar test, depending on variable distribution.
*Holm-Bonferroni correction. DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; M, missing; SBP, systolic blood pressure.
Table 2.
Changes in metabolic and insulin requirements
| Parameter in which change was measured | All patients | SG group | RYGB group | P value* |
|---|---|---|---|---|
| TWL (%) | 29.7 (29.4, 30.3) | 28.6 (28.5, 28.7) | 31.6 (31.4, 31.9) | 0.346 |
| BMI (kg/m²) | 11.6 (11.5, 11.7) | 11.4 (11.1, 11.4) | 12.6 (12.5, 12.8) | 0.346 |
| HbA1c (mmol/mol) | 4.0 (3.0, 5.0) | 2.0 (0.7, 3.6) | 5.7 (4.0, 7.8) | 0.118 |
| Insulin total | ||||
| Units/day | 45.0 (44.0, 48.3) | 42.0 (38.7, 44.0) | 56.4 (51.5, 60.5) | 0.087 |
| Units/kg/day | 0.42 (0.40, 0.44) | 0.39 (0.37, 0.41) | 0.49 (0.44, 0.50) | 0.118 |
| Units/day, % | 58.2 (56.7, 59.8) | 53.5 (51.9, 57.0) | 63.0 (61.0, 63.6) | 0.182 |
| Cholesterol | 0.50 (0.45, 0.60) | 0.25 (0.13, 0.37) | 1.16 (0.96, 1.29) | <0.001 |
| LDL | 0.44 (0.40, 0.52) | 0.33 (0.16, 0.42) | 0.70 (0.54, 0.83) | 0.014 |
| HDL | −0.19 (−0.21 to −0.15) | −0.24 (−0.28 to −0.18) | −0.11 (−0.18 to −0.09) | 0.118 |
| Triglycerides | 0.50 (0.44, 0.57) | 0.50 (0.41, 0.63) | 0.50 (0.32, 0.64) | 0.675 |
Values are presented as median (IQR) for nonnormally distributed variables and were compared using the Mann-Whitney U test. *Holm–Bonferroni correction.
Anthropometric Changes
Preoperative BMI was 40.2 ± 5.6 kg/m2, which decreased to 28.5 ± 4.8 kg/m2 at 12 months (P < 0.001) (Table 1). Waist circumference declined from 120.6 ± 16.3 cm to 99.5 ± 13.0 cm (P < 0.001). Individuals living with obesity and T1D achieved substantial weight loss, with a median TWL percentage of 29.7% (IQR = 29.4, 30.3) at the end of the follow-up (Fig. 1 and Table 2).
Figure 1.
Changes in TWL, HbA1c, and insulin over time from surgery and up to 1 year after. Solid lines represent the mean, the shaded area indicates the 95% CI.
Glycemic Outcomes
Mean HbA1c decreased from 64.0 (IQR = 57.0, 74.0) mmol/mol preoperatively to 60.0 (IQR = 53.8, 68.0) mmol/mol at 12 months after MBS (P < 0.001) (Table 1 and Fig. 1). Compared with their baseline values, total daily insulin use and weight-adjusted total daily insulin use decreased after 12 months (respectively, 82.6 [IQR = 60.0, 115.3] units/day vs. 36.3 [IQR = 24.0, 50.1] units/day and 0.75 [IQR = 0.58, 1.00] units/kg/day vs. 0.32 [IQR = 0.23, 0.43] units/kg/day) (both P < 0.001).
Lipid Profile and Renal Function
All cardiometabolic parameters improved significantly after surgery (all P < 0.001), with decreases in LDL, total cholesterol, triglycerides, and blood pressure measurements, and an increase in HDL level (Table 1). Renal function remained largely unchanged after surgery.
Insulin Delivery and Procedure Type
Subgroup analyses (Supplementary Table 3) showed that patients treated with continuous subcutaneous insulin infusion (CSII) had slightly higher body weight, larger waist circumference, and higher blood pressure at baseline. Patients receiving MDIs and those receiving CSII improved in weight, BMI, HbA1c, insulin dose, and lipid profiles at 12 months. RYGB resulted in greater cholesterol reduction than SG (respective changes from baseline: 1.16 vs. 0.25 mmol/L; P < 0.001) (Table 2).
GAMs and LMMs
GAMs revealed a nonlinear trajectory of weight loss, with the greatest reduction occurring shortly after MBS and then a slower subsequent decline (Supplementary Table 4 and Fig. 2). Women lost more weight than men at 12 months (β = 1.97 [95% CI 0.24–3.69]; P = 0.026). Higher TWL was associated with lower weight-adjusted insulin requirements (88% variance explained).
Figure 2.
Evolution of total insulin (A) and TWL (B) over 12 months after MBS. C: Relationship between TWL and percentage change in insulin at the first visit after MBS.
Weight-adjusted total daily insulin also decreased nonlinearly, with every 10% TWL corresponding to a 0.10 units/kg/day reduction (β = −0.01 [95% CI −0.01 to0.00], P = 0.010; 63% variance explained) (Supplementary Table 5). The most pronounced reduction in insulin requirements occurred during the first months postoperatively (Fig. 2). An LMM examining the change from preoperative baseline to the first available postoperative measurement (median 3.45 months; IQR = 2.65–8.5) showed that greater TWL was associated with a larger percentage reduction in insulin use (β = 2.74 [95% CI 2.34–3.14]; P < 0.001), and the rate of reduction slowed over time (β = −2.59 [95% CI −3.71 to −1.47]; P < 0.001) (Supplementary Table 6).
Higher baseline HbA1c values predicted higher postoperative HbA1c (P < 0.001) and a significant interaction with weight-adjusted insulin dose (P = 0.005) (Supplementary Table 7), meaning that the effect of insulin use on postoperative HbA1c differed depending on the baseline HbA1c level.
Sensitivity analyses in confirmed and unconfirmed cases were comparable to the total population analyses (Supplementary Tables 8–15).
Complications
Within 30 days after surgery, eight cases of diabetic ketoacidosis, nine cases of hypoglycemia, and one case of gastric ulcer-related bleeding were reported among the 162 participants. More details are provided in the Supplementary Methods.
Conclusions
To our knowledge, this is the largest multicenter study to date of MBS in adults with T1D (9,10), comparable in size to previous meta-analyses (11,12). Our findings show that MBS was associated with substantial metabolic improvements, including significant weight loss and favorable changes in lipid profiles 12 months after surgery. Longitudinal modeling showed that baseline HbA1c was the strongest predictor of postoperative glycemic control, and individuals with both high preoperative HbA1c and higher insulin needs benefited the most. Weight loss did not differ significantly between RYGB and SG or between MDI and CSII users.
Moreover, insulin requirements decreased substantially, and greater weight loss was associated with a more pronounced reduction in insulin needs, reflecting improved insulin sensitivity. The steepest decline in insulin use was observed in the early postoperative period. Based on our model, an average postoperative weight loss of 17–18% at around 3 months corresponds to an estimated 47–49% reduction in insulin requirements, comparable to what has been previously shown (10,13,14). These findings underscore the importance of close postoperative monitoring and proactive insulin adjustment to prevent hypoglycemia. Nevertheless, HbA1c declined modestly, by 4.0 mmol/mol on average, remaining above recommended targets and in line with previous literature reporting minor or no changes in HbA1c after MBS (11,13,14).
However, pharmacotherapy may be appropriate for individuals with lower BMI; it offers moderate weight loss and glycemic benefits with a favorable safety profile. In studies, pharmacotherapy, including GLP-1 receptor agonists such as liraglutide and semaglutide, induced modest weight loss (2–5 kg over 26 weeks), improved glycemic control, and reduced insulin requirements (15,16). These studies mainly included individuals with BMI in the overweight to obesity class I range (26–35 kg/m2), and adverse events were generally mild (15,16). In contrast, our cohort undergoing MBS had a much higher baseline BMI (mean ± SD, 40.2 ± 5.6 kg/m2) and experienced substantially greater weight loss, significant reductions in insulin requirements, and improvements in cardiometabolic parameters. MBS provides greater metabolic benefits for those with higher BMI, though surgical risks must be considered. Treatment decisions should be individualized, considering BMI, cardiometabolic risk, patient preference, and expected benefit.
This multicenter study with a large sample size represents one of the largest cohorts evaluating MBS in people with T1D and obesity. Its longitudinal data offer insight into how MBS effects evolve during the first postoperative year. However, several limitations should be noted. The retrospective design introduces potential bias, and the absence of a nonsurgical control group limits comparisons. Follow-up measurements were not standardized, and some data were missing. CGM and self-monitored glucose data were not routinely available, restricting assessment of glycemic variability beyond HbA1c. Although follow-up was adequate for short-term outcomes, longer-term data are needed to assess durability and late complications. Data of postoperative complications, hospitalizations, and key clinical outcomes (e.g., severe hypoglycemia, nutritional deficiencies, gastrointestinal symptoms) were not consistently collected, nor were data on functional outcomes or quality of life. These omissions are major limitations. Future prospective studies should incorporate standardized follow-up, comprehensive complication monitoring, patient-reported outcomes, and additional metabolic measures such as insulin resistance.
In conclusion, MBS is associated with significant weight loss, lower insulin requirements, and improved lipid profile, potentially reducing the risk for cardiovascular disease. These benefits were observed across the cohort, irrespective of insulin delivery method.
This article contains supplementary material online at https://doi.org/10.2337/figshare.31136113.
Article Information
Acknowledgments. The authors thank Carlos Marin (Clinical and Experimental Endocrinology, Department of Chronic Diseases and Metabolism, KU Leuven, Leuven, Belgium) for providing medical writing support and editorial assistance. This work is part of the Stratification of Obesity Phenotypes to Optimize Future Obesity Therapy (SOPHIA) project (www.imisophia.eu).
This article reflects only the authors’ views. Neither the IMI, the European Union, EFPIA, nor any associated partners are responsible for the use of the information it contains. Medtronic acted solely as a funder and was not involved in the research or in the evaluation or editing of the scientific content of this article.
Duality of Interest. C.M. serves or has served on advisory panels for Novo Nordisk, Sanofi, Eli Lilly and Company, Novartis, Boehringer Ingelheim, Bayer, Roche, Medtronic, Imcyse, Insulet, Biomea Fusion, and Vertex. Financial compensation for these activities by C.M. has been received by KU Leuven, which has received research support for C.M. from Medtronic, Imcyse, Novo Nordisk, Sanofi, and ActoBio Therapeutics; C.M. serves or has served on speakers bureaus for Novo Nordisk, Sanofi, Eli Lilly and Company, Medtronic, and Boehringer Ingelheim. Financial compensation for these activities has been received by KU Leuven. C.M. is president of the European Association for the Study of Diabetes (EASD). All external support of EASD can be found at www.easd.org. R.v.d.M. is employed by NOK and NOK received funding from Medtronic for conducting this study. No other potential conflicts of interest relevant to this article were reported.
Author Contributions. R.v.d.M. and S.P. analyzed data and wrote and edited the manuscript. N.S., A.v.L., J.R., C.H.d.P., C.W.R., R.H., C.M., R.L., and B.V.d.S. contributed to discussion and reviewed and edited the manuscript. M.S., L.T., M.L., E.D., B.D., Y.v.N., A.v.B., E.H., E.A.O., J.A.K. and M.I. researched data and reviewed and edited the manuscript. B.V.d.S. and R.v.d.M. are guarantors of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were Elizabeth Selvin and Matthew Crowley.
Funding Statement
SOPHIA has received funding from the Innovative Medicines Initiative 2 Joint Undertaking (grant 875534), supported by the European Union’s Horizon 2020 research and innovation program and the European Federation of Pharmaceutical Industries and Associations (EFPIA), with additional support from T1D Exchange, Breakthrough T1D (formerly JDRF), and Obesity Action Coalition. The Nederlandse Obesitas Kliniek (NOK) received funding from Medtronic for the conduct of this study.
Supporting information
References
- 1. van Laar ADE, Gallardo-Nuell L, Lucio SP, et al. Trends in baseline HbA1c and body-mass index in randomised trials of people with type 1 diabetes from 1993 to 2025: an IMI2 SOPHIA systematic review and meta-analysis. EClinicalMedicine 2025;89:103589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Al Ozairi E, Steenackers N, Pazmino S, et al. Prevalence of obesity in people with and without type 1 diabetes across Belgium, Kuwait, and Mexico: an IMI2 SOPHIA study. EClinicalMedicine 2024;77:102869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Steenackers N, Feldman AN, Mathieu C, et al. The double burden: navigating type 1 diabetes and obesity. Clin Obes 2024;14:e12645. [DOI] [PubMed] [Google Scholar]
- 4. Thaher O, Iaroshevych V, Driouch J, Hukauf M, Croner RS, Stroh C. Current status of metabolic surgery in patients with type I diabetes mellitus and obesity: a nationwide multicenter study. Langenbecks Arch Surg 2023;408:46. [DOI] [PubMed] [Google Scholar]
- 5. Chow A, Switzer NJ, Dang J, et al. A systematic review and meta-analysis of outcomes for type 1 diabetes after bariatric surgery. J Obes 2016;2016:6170719. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Brethauer SA, Aminian A, Rosenthal RJ, Kirwan JP, Kashyap SR, Schauer PR. Bariatric surgery improves the metabolic profile of morbidly obese patients with type 1 diabetes. Diabetes Care 2014;37:e51-52–e52 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Dessify B, Wood C, Parker D, Carmichael D, Petrick A, Daouadi M. Is there a role for bariatric surgery in patients with severe obesity and type 1 diabetes? Surg Obes Relat Dis 2022;18:177–181 [DOI] [PubMed] [Google Scholar]
- 8. Lannoo M, Dillemans B, Van Nieuwenhove Y, et al. Bariatric surgery induces weight loss but does not improve glycemic control in patients with type 1 diabetes. Diabetes Care 2014;37:e173–e174 [DOI] [PubMed] [Google Scholar]
- 9. Mohammed N, Buckley A, Elsheikh M, et al. Bariatric surgery in the treatment of patients with obesity and type 1 diabetes: a retrospective study of clinical data. Diabetes Obes Metab 2021;23:1562–1570 [DOI] [PubMed] [Google Scholar]
- 10. Vilarrasa N, San Jose P, Rubio MÁ, Lecube A. Obesity in patients with type 1 diabetes: links, risks and management challenges. Diabetes Metab Syndr Obes 2021;14:2807–2827 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Ashrafian H, Harling L, Toma T, et al. Type 1 diabetes mellitus and bariatric surgery: a systematic review and meta-analysis. Obes Surg 2016;26:1697–1704 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Mahawar KK, De Alwis N, Carr WRJ, Jennings N, Schroeder N, Small PK. Bariatric surgery in type 1 diabetes mellitus: a systematic review. Obes Surg 2016;26:196–204 [DOI] [PubMed] [Google Scholar]
- 13. Moreno-Fernandez J, Chico A, Working Group of Diabetes and Technology of the Spanish Diabetes Association, Spain . Bariatric surgery results in patients with type 1 diabetes mellitus on continuous subcutaneous insulin infusion therapy. Endocrinol Nutr 2016;63:571–572 [DOI] [PubMed] [Google Scholar]
- 14. Middelbeek RJW, James-Todd T, Patti M-E, Brown FM. Short-term insulin requirements following gastric bypass surgery in severely obese women with type 1 diabetes. Obes Surg 2014;24:1442–1446 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Shah VN, Akturk HK, Kruger D, et al. Semaglutide in adults with type 1 diabetes and obesity. NEJM Evid 2025;4:EVIDoa2500173. [DOI] [PubMed] [Google Scholar]
- 16. Park J, Ntelis S, Yunasan E, et al. Glucagon-like peptide 1 analogues as adjunctive therapy for patients with type 1 diabetes: an updated systematic review and meta-analysis. J Clin Endocrinol Metab 2023;109:279–292 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.