Bariatric metabolic surgery (BMS) was associated with lower all-cause mortality compared with glucagon-like peptide-1 receptor agonists (GLP-1RAs) treatment in individuals with a diabetes duration of ten years or less and without a prior history of cardiovascular disease (CVD) or congestive heart failure (CHF), over a median follow-up of 6.8 years.
According to Israeli investigators who carried out the research, weight loss was identified as a mediating factor in this association. However, there was no difference between BMS and treatment with GLP-1RAs was observed in the risk of mortality among individuals with a longer duration of diabetes (>10 years) or in the risk of MACEs among all patients.
For the study, patients who underwent BMS or were treated with GLP-1RAs between January 2008 and December 2021, were matched and followed up for mortality and MACEs. Surgical patients were defined as having undergone laparoscopic banding, Roux-en-Y gastric bypass or laparoscopic sleeve gastrectomy during these years. Patients were considered as treated with GLP-1RAs if they purchased first-generation GLP-1RAs (liraglutide, dulaglutide, exenatide, lixisenatide, or insulin degludec) for at least six months within a period of 12 consecutive months from 2008 to 2021.
Eligible patients were matched 1:1 (1 patient who underwent BMS to 1 patient treated with GLP-1RAs) based on the following variables: sex, age group (in 5-year increments), BMI (in 5-unit increments), index date (in 2-year increments), and diabetes duration (the time from diagnosis until the index date, in 5-year increments).
The primary outcome, all-cause mortality, was obtained from Ministry of Interior data. Secondary outcomes included nonfatal MACEs (myocardial infraction, stroke, or ischemic heart disease or the performance of percutaneous transluminal coronary angioplasty or coronary artery bypass graft during the follow-up period) and maximal and long-term changes in haemoglobin A1c (HbA1c) concentration and in BMI. Long-term change refers to the latest time point available.
Outcomes
In total, 3,035 matched pairs (6070 individuals) were included in the analyses (mean [SD] age, 51.0 [9.5] years; 3,938 women [64.9%] and 2,132 men [35.1%]). The mean BMI at the index date was 41.3. Patients who underwent BMS were less likely to have hyperlipidaemia and hypertension, and used less insulin and fewer agents acting on the renin-angiotensin system, sodium-glucose cotransporter-2 inhibitors, and other blood glucose–lowering drugs, compared with individuals who were treated with GLP-1RAs.
The median follow-up time was 6.8 years and the maximum follow-up time was 12 years. A change in treatment, resulting in discontinuation of follow-up, was observed among 352 of 3035 patients who underwent BMS (11.6%) and who had also initiated treatment with GLP-1RAs during the follow-up period. In addition, 252 of 3035 patients taking GLP-1RAs (8.3%) underwent BMS during follow-up.
The duration of diabetes was shown to interact with mortality associated with GLP-1RA vs BMS (p=0.03 for the interaction treatment type × diabetes duration). Therefore, all results are reported separately for patients with a diabetes duration of ten years or less or more than ten years at index date.
The cohort comprised 2,371 pairs of patients with a diabetes duration of ten years or less and 664 pairs of patients with a diabetes duration of longer than ten years. No interaction was observed between age at baseline and type of treatment in their association with mortality.
In patients with a diabetes duration of ten years or less, 42 of those who underwent BMS (266 per 100 000 person-years) and 119 of those treated with GLP-1RAs (785 per 100 000 person-years) died during the follow-up period. The Kaplan-Meier estimated mortality curves demonstrated significantly lower mortality among the patients who underwent BMS than among those who were treated with GLP-1RAs (p<0.001 by log-rank test).
The adjusted Cox proportional hazards regression model demonstrated a lower risk for death among those who underwent BMS compared with those who were treated with GLP-1RAs (hazard ratio [HR], 0.38; 95% CI, 0.25-0.58). The association was preserved when the maximal change in HbA1c concentration during follow-up was included in the model (HR, 0.43; 95% CI, 0.27-0.99) but disappeared once the percentages of the maximal change in BMI level were also introduced into the model (HR, 0.79; 95% CI, 0.43-1.48). The mediation analysis with weight loss as a mediator revealed a significant association between treatment type and weight loss (estimate, −18.42; p<0.01), as well as between weight loss and mortality (HR, 1.05; 95% CI, 1.02-1.08).
Similar results were observed when individuals who received a diagnosis of cancer in the first 2 years of follow-up were excluded from the analysis. Among the patients with a diabetes duration of longer than 10 years, 32 of those who underwent BMS (695 per 100 000 person-years) and 56 of those treated with GLP-1RAs (1139 per 100 000 person-years) died during the follow-up period. The Kaplan-Meier estimated mortality curves demonstrated significantly lower mortality among the patients who underwent BMS compared with those who were treated with GLP-1RAs (p=0.047 by log-rank test). However, the adjusted Cox proportional hazards regression model demonstrated no statistically significant difference between groups (HR, 0.65; 95% CI, 0.39-1.08).
“In this retrospective cohort study, among individuals with obesity and a diabetes duration of 10 years or less, but with no prior history of ischemic heart disease, ischemic stroke, or congestive heart failure, BMS was associated with a 62% reduction (HR, 0.38; 95% CI, 0.25-0.58) in mortality compared with GLP-1RAs, after adjustment for potential confounders,” the authors write. “…In this study, BMS did not show a survival advantage compared with treatment with GLP-1RAs among patients with a diabetes duration of longer than 10 years, despite the long-term decrease in BMI. This finding may be explained by the adverse effects of prolonged diabetes duration, which masked the benefit associated with weight reduction.”
The findings were reported in the paper, ‘Bariatric Metabolic Surgery vs Glucagon-Like Peptide-1 Receptor Agonists and Mortality’, published in JAMA Network Open. To access this paper, please click here
Comments