Abstract
Background and Aims
Glucagon-like peptide-1 receptor agonists (GLP-1RA) are increasingly used in adults with type 2 diabetes (T2D), with or without obesity. The incidence of gastrointestinal (GI) adverse effects (AEs) of GLP-1RA in T2D is unclear. This study aimed to evaluate all-cause mortality and GI AEs in T2D patients treated with GLP-1RA compared to those treated with sodium–glucose cotransporter-2 inhibitors (SGLT-2i).
Methods
A retrospective cohort study used electronic health records from the TriNetX Multi-Institutional Database (January 1, 2021, to December 31, 2022). T2D patients on GLP-1RA were compared to those on SGLT-2i. Primary outcomes were incident GI AEs and all-cause mortality. 1:1 propensity score matching was performed to reduce confounding, and adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were calculated for each outcome.
Results
The study included 3.2 million adults with T2D, with 104,947 prescribed a GLP-1RA compared with a matched cohort prescribed a SGLT-2i. After matching, baseline characteristics were similar, with a mean age of 62 ± 12 years and mean glycated hemoglobin of 8.0 ± 2.0%. About 30% of participants were from under-represented minority groups. At 24-month follow-up, patients prescribed a GLP-1RA had higher odds of being diagnosed with gastroparesis (GP) and gastroesophageal reflux disease (GERD) compared those prescribed a SGLT-2i, with aORs of 1.24 (95% CI, 1.11–1.38) for GP and 1.14 (95% CI, 1.11–1.18) for GERD. The risk of acute pancreatitis was lower in the GLP-1RA group. All-cause mortality at 24 months had an odds ratio of 0.83 (95% CI: 0.80, 0.87) compared to SGLT-2i.
Conclusion
After 24 months of follow-up, patients treated with a GLP-1RA had higher odds of GP and GERD compared to those treated with a SGLT-2i. However, the odds of acute pancreatitis were lower in the GLP-1RA group. All-cause mortality was reduced by 17% in the GLP-1RA group.
Keywords: GLP-1 receptor agonists, Type 2 diabetes, Gastrointestinal adverse effects, Gastroparesis, SGLT-2 inhibitor
Background and Aims
Diabetes mellitus (DM) is a global health concern, affecting approximately 10.5% of adults worldwide, with projections indicating a significant increase to 12.2% by 2045.1 In the United States alone, an estimated 11.3% of the population (38.4 million people), have DM, primarily type 2 diabetes (T2D), which accounts for over 95% of DM cases.2, 3, 4, 5 Cardiovascular disease (CVD) and stroke are the leading causes of mortality among individuals with T2D. People with T2D have an increased CVD risk, with CVD occurring 2–4 times more frequently and resulting in a threefold increase in mortality compared to the general population.6
The GLP-1RA and dual-agonists (GLP-1RA + glucose dependent insulinotropic peptide) are recommended therapies for T2D, particularly in obese individuals associated with comorbidities. GLP-1RA stimulate glucose-dependent insulin release from pancreatic beta cells, suppress the paradoxical increase of glucagon after meals, and delay gastric emptying, thereby contributing to improved glycemic control and weight loss. These agents not only improve glycemic control but also confer cardiovascular (CVD) and renal (diabetic kidney disease) benefits as documented by several randomized controlled trials.6, 7, 8, 9, 10, 11 SGLT-2i are widely used in T2D as they also improve CVD and diabetic kidney disease outcomes12 but are not known to have significant gastrointestinal (GI) effects. Common AEs from SGLT-2i include metabolic and genitourinary issues. Hence we decided to compare the benefits and GI adverse effects (AEs) of these 2 classes of medications frequently prescribed in T2D.
Despite the CVD benefits associated with GLP-1RA, the comprehensive safety profile, particularly GI AEs remains incompletely characterized, especially in real-world populations.13, 14, 15, 16 Common GI AEs include gastroparesis (GP), gastroesophageal reflux disease (GERD), ileus, intestinal obstruction, pancreatitis, and biliary complications13,17, 18, 19, 20, 21, 22, 23, 24 though the risk of GLP-1RA in T2D in large studies may be different than in other conditions or smaller studies. The incidence and risk of these events may also vary based on population characteristics, drug formulations, and study designs, which are underexplored in large, diverse, real-world cohorts.13, 14, 15, 16 We evaluated all-cause mortality and GI AEs above in adults with T2D that were prescribed GLP-1RA or dual-agonists vs SGLT-2i over 2 years using a large multi-institutional deidentified database (TriNetX).
Methods
Database
We utilized deidentified electronic medical records from the large US Collaborative Network comprising 64 health-care organizations (HCOs) within the TriNetX Research Network, encompassing a total of 113 million patients. The dataset included demographic information, diagnoses (coded according to the International Classification of Diseases, Tenth Revision, ICD-10), procedures (coded in ICD-10-procedural coding system or Current Procedural Terminology), medication records, and laboratory tests (coded in LOINC). Additional ICD codes used in this data analysis are provided in the Appendix.
TriNetX is a health collaborative clinical research platform that gathers real-time Electronic Medtronic Records from various HCOs. TriNetX adheres to Health Insurance Portability & Accountability Act compliance and is exempt from institutional review board oversight as only deidentified data are used for analyses.
Patient Selection
Between January 1, 2021, and December 30, 2022, we identified a total of 3.2 million patients aged 18 years and older with a diagnosis of T2D (ICD-10 code E11). The cohort included 30% under-represented minorities (9% Hispanic, 16.8% African American, and 4.2% Asian). Patients were categorized into 2 groups before propensity matching1: Patients prescribed a GLP-1RA (dulaglutide, semaglutide, liraglutide, or tirzepatide) during the study period, with no concurrent prescription for an SGLT-2i, n = 194,669, and2 patients prescribed a SGLT-2i (dapagliflozin, empagliflozin, and canagliflozin) during the study period, with no concurrent prescription for a GLP-1RA, n = 140,934 (Figure 1).
Figure 1.
Patient selection flow diagram. The last 2 columns represent the final numbers of subjects in the 2 groups after propensity score matching. Patients with a history of bariatric surgery, gastric surgery, fundoplication, inflammatory bowel disease, per oral endoscopic myotomy, and pyloroplasty were excluded from the study. Patients with at least 2 ambulatory visits in the chart were included to ensure adequate follow-up.
After propensity matching there were 104,947 patients in both treatment groups (the GLP-1RA group and the SGLT-2i group) (Table 1 and Figure 1). We evaluated GI AEs such as GP, GERD, ileus, intestinal obstruction, acute pancreatitis, acute cholecystitis, cholelithiasis, and cholangitis at 3, 6, 12, and 24 months following the index date (defined as prescription of GLP-1RA and prescription of SGLT-2i). We also assessed all-cause mortality and incident major adverse cardiovascular events (MACEs), including myocardial infarction (MI), stroke, at 3, 6, 12, and 24 months following the index date. Patients with a prior history or diagnosis of these outcomes (GI AEs) at any time before the index date were excluded from the study.
Table 1.
Baseline Characteristics of T2D on GLP-1RA vs SGLT-2i (After Propensity Score Matching)
| T2D + GLP-1RA N = 104,947 |
T2D + SGLT-2i N = 104,947 |
P value | Std Diff | |
|---|---|---|---|---|
| Demographics | ||||
| Age at Index | 62.0 ± 12.0 | 62.0 ± 12.0 | .93 | <0.001 |
| White | 63,428 (60.4%) | 63,221 (60.2%) | .36 | 0.004 |
| Female | 44,427 (42.3%) | 44,899 (42.8%) | .04 | 0.009 |
| Unknown ethnicity | 21,752 (20.7%) | 21,557 (20.5%) | .29 | 0.005 |
| Not Hispanic or Latino | 73,712 (70.2%) | 74,085 (70.6%) | .07 | 0.008 |
| Hispanic or Latino | 9483 (9.0%) | 9305 (8.9%) | .17 | 0.006 |
| Black or African American | 17,591 (16.8%) | 17,795 (17,795) | .23 | 0.005 |
| Male | 55,922 (53.3%) | 55,395 (52.8%) | .02 | 0.01 |
| Asian | 4418 (4.2%) | 4486 (4.3%) | .46 | 0.003 |
| Comorbidities | ||||
| Parkinson disease | 636 (0.6%) | 642 (0.6%) | .87 | 0.001 |
| Heart Failure | 16,088 (15.3%) | 16,477 (15.7%) | .02 | 0.01 |
| Ischemic heart disease | 28,155 (26.8%) | 28,352 (27.0%) | .33 | 0.004 |
| Chronic kidney disease | 19,829 (18.9%) | 19,944 (19.0%) | .52 | 0.003 |
| Diseases of liver | 14,828 (14.1%) | 14,872 (14.2%) | .78 | 0.001 |
| Alcohol-related disorders | 3842 (3.7%) | 3869 (3.7%) | .75 | 0.001 |
| Nicotine dependence | 15,235 (14.5%) | 15,377 (14.7%) | .38 | 0.004 |
| Constipation | 12,388 (11.8%) | 12,471 (11.9%) | .57 | 0.002 |
| Polyneuropathies | 9801 (9.3%) | 9746 (9.3%) | .68 | 0.002 |
| Neoplasms | 33,642 (32.1%) | 33,630 (32.0%) | .95 | <0.001 |
| Disorders of lipoprotein metabolism and other lipidemias | 72,575 (69.2%) | 73,044 (69.6%) | .03 | 0.01 |
| Hypertensive disease | 75,116 (71.6%) | 75,562 (72.0%) | .03 | 0.009 |
| Medications | ||||
| Opioids | 52,651 (50.2%) | 52,997 (50.5%) | .13 | 0.007 |
| insulin | 46,296 (44.1%) | 46,383 (44.2%) | .7 | 0.002 |
| Laxatives | 44,666 (42.6%) | 45,020 (42.9%) | .12 | 0.007 |
| Proton pump inhibitors | 39,242 (37.4%) | 39,479 (37.6%) | .28 | 0.005 |
| Aspirin | 41,303 (39.4%) | 41,778 (39.8%) | .03 | 0.009 |
| Beta blocking agents | 47,923 (45.7%) | 48,277 (46.0%) | .12 | 0.007 |
| Antilipidemic agents | 72,752 (69.3%) | 72,989 (69.5%) | .26 | 0.005 |
| Angiotensin-converting enzyme inhibitors | 45,081 (43.0%) | 45,256 (43.1%) | .44 | 0.003 |
| Angiotensin receptor blocker | 32,498 (31.0%) | 32,557 (31.0%) | .78 | 0.001 |
| BMI/HbA1c | ||||
| BMI | 34.1 ± 7.8 | 33.0 ± 7.3 | <.001 | 0.149 |
| HbA1c | 8.0 ± 2.0 | 8.0 ± 1.9 | <.001 | 0.021 |
Plus -minus values are means ± standard deviation. Numbers in parentheses are percentages. After propensity matching baseline characteristics were well balanced.
Patients with a history of bariatric surgery, gastric surgery, fundoplication, inflammatory bowel disease, per oral endoscopic myotomy, and pyloroplasty were excluded from the study since these diagnoses can affect gastrointestinal motility and may mask or amplify the effects of GLP-1RA. Patients with at least 2 ambulatory visits in the chart were included to ensure adequate follow-up.
Statistical Analysis
TriNetX compares baseline characteristics using the chi-square test for categorical variables and the t-test for continuous variables. To mitigate the influence of confounding factors, we employed propensity score matching to create groups with similar baseline characteristics. We utilized TriNetX's built-in function to match the 2 groups at a 1:1 ratio through nearest-neighbor matching. The TriNetX platform uses input matrices with user-selected covariates to conduct logistic regression analysis, generating propensity scores for each individual subject. These propensity scores are then applied to match patients using a greedy nearest-neighbor algorithm with a caliper width of 0.1 pooled standard deviations. To minimize bias introduced by the order of data, TriNetX randomizes the row sequence before matching. Postmatching P values are provided to assess covariate balance between the 2 cohorts.
Propensity score matching was conducted for variables including age, sex, race, ethnicity, chronic kidney disease, hyperlipidemia, hypertension, liver disease, neoplasms, ischemic heart disease heart failure, alcohol-related disorders, nicotine dependence, constipation, Parkinson’s disease, neuropathy, and medications such as insulin, aspirin, beta blockers, antilipemic agents, angiotensin converting enzyme inhibitors, angiotensin receptor blockers, proton pump inhibitors, laxatives, opioids as well as laboratory values such as glycated hemoglobin (HbA1c), in addition to body mass index (BMI). Variables used for propensity score matching were obtained from any time before the index date. For continuous variables such as age, BMI, and HbA1c, the most recent value recorded before the index date was used.
Following propensity score matching, each outcome was evaluated and reported as an adjusted odds ratio (aOR) with a 95% confidence interval (CI). Statistical significance was defined as a 2-sided P value < .05. All analyses were conducted using the TriNetX Analytics function.
Results
Demographics
Following propensity score matching, minimal differences were observed in demographics, comorbidities, medications, and laboratory findings, indicating well-balanced matching (Table 1).
Patients prescribed a GLP-1RA had higher mean BMI compared to those prescribed a SGLT-2i (Table 2). The mean HbA1c remained elevated (>7.0%) in both groups. However, mean HbA1c tended to be lower at 24 months in patients prescribed a GLP-1RA treated group compared to patients prescribed a SGLT-2i (Table 2).
Table 2.
Mean HbA1c and BMI in Patients With T2D on GLP-1RA vs SGLT-2i
| Time since index date (mo) | Hemoglobin A1c (%) |
BMI (kg/m2) |
||||
|---|---|---|---|---|---|---|
| Mean HbA1c in T2D + GLP-1RA | Mean HbA1c in T2D + SGLT-2i | P value | Mean BMI in T2D + GLP-1RA | Mean BMI in T2D + SGLT-2i | P value | |
| 3 | 7.54 ± (SD 1.66) | 7.65 ± (SD 1.60) | <.001 | 33.80 ± (SD 7.70) | 32.33 ± (SD 7.24) | <.001 |
| 6 | 7.35 ± (SD 1.59) | 7.53 ± (SD 1.52) | <.001 | 33.56 ± (SD 7.62) | 32.18 ± (SD 7.13) | <.001 |
| 12 | 7.29 ± (SD 1.63) | 7.49 ± (SD 1.54) | <.001 | 33.33 ± (SD 7.58) | 32.04 ± (SD 7.10) | <.001 |
| 24 | 7.29 ± (SD 1.68) | 7.51 ± (SD 1.58) | <.001 | 33.06 ± (SD 7.56) | 31.88 ± (SD 7.10) | <.001 |
Plus -minus values are means ± standard deviation. BMI values are presented as kg/m2. Patients on GLP-1RA had higher mean BMI compared to the SGLT-2i group throughout the study duration. The mean HbA1c remained elevated (>7.0%) in both groups across all time points indicating suboptimal glucose control. However, mean HbA1c was lower at all times in the GLP-1RA treated group.
GI AEs
Patients with T2D prescribed a GLP-1RA had a higher risk of receiving a diagnosis of GP and GERD at 24 months compared to patients prescribed a SGLT2i (aOR 1.24, 95% CI: 1.11–1.38 and aOR 1.14, 95% CI: 1.11–1.18, respectively). In contrast, the risk of ileus and acute pancreatitis was lower in patients prescribed a GLP-1RA compared to patients prescribed a SGLT2i (aOR 0.80, 95% CI: 0.70–0.91 and aOR 0.76, 95% CI: 0.67–0.85 respectively). No significant differences in the risk of acute cholecystitis, cholelithiasis, cholangitis or intestinal obstruction were observed at any time points (Table 3).
Table 3.
Gastrointestinal Outcomes in Patients With T2D on GLP-1RA vs SGLT-2i (After Propensity Score Matching)
| Time since index date (m) | T2D + GLP-1RA (n) |
T2D + SGLT-2i (n) |
aOR | 95% CI |
|---|---|---|---|---|
| GPa | ||||
| 3 | 145 | 138 | 1.05 | (0.83–1.32) |
| 6 | 266 | 227 | 1.17 | (0.98–1.40) |
| 12 | 447 | 400 | 1.12 | (0.98–1.28) |
| 24 | 723 | 583 | 1.24 | (1.11–1.38) |
| Ileus | ||||
| 3 | 91 | 131 | 0.69 | (0.53–0.91) |
| 6 | 143 | 202 | 0.71 | (0.57–0.88) |
| 12 | 259 | 326 | 0.79 | (0.67–0.93) |
| 24 | 400 | 502 | 0.80 | (0.70–0.91) |
| GERDa | ||||
| 3 | 2097 | 2247 | 0.96 | (0.90–1.02) |
| 6 | 3466 | 3465 | 1.03 | (0.98–1.08) |
| 12 | 5377 | 5142 | 1.08 | (1.04–1.13) |
| 24 | 7911 | 7210 | 1.14 | (1.11–1.18) |
| Intestinal obstruction | ||||
| 3 | 85 | 78 | 1.09 | (0.80–1.48) |
| 6 | 136 | 149 | 0.91 | (0.72–1.15) |
| 12 | 228 | 243 | 0.94 | (0.78–1.12) |
| 24 | 366 | 381 | 0.96 | (0.83–1.11) |
| Acute pancreatitis | ||||
| 3 | 98 | 165 | 0.58 | (0.46–0.75) |
| 6 | 166 | 267 | 0.61 | (0.50–0.74) |
| 12 | 309 | 420 | 0.72 | (0.63–0.84) |
| 24 | 496 | 645 | 0.76 | (0.67–0.85) |
| Acute cholecystitis | ||||
| 3 | 56 | 49 | 1.14 | (0.78–1.68) |
| 6 | 103 | 90 | 1.14 | (0.86–1.52) |
| 12 | 179 | 161 | 1.11 | (0.90–1.38) |
| 24 | 305 | 279 | 1.09 | (0.93–1.29) |
| Cholelithiasis | ||||
| 3 | 397 | 428 | 0.93 | (0.81, 1.06) |
| 6 | 662 | 680 | 0.97 | (0.87, 1.08) |
| 12 | 1153 | 1164 | 0.99 | (0.91, 1.07) |
| 24 | 1881 | 1886 | 1.00 | (0.93, 1.06) |
| Cholangitis | ||||
| 3 | 33 | 26 | 1.27 | (0.76, 2.12) |
| 6 | 54 | 46 | 1.17 | (0.79, 1.74) |
| 12 | 94 | 94 | 1.00 | (0.75, 1.33) |
| 24 | 137 | 139 | 0.99 | (0.78, 1.25) |
N = number of subjects with the respective event. 95% CI are provided for each column. AEs were selected based on the safety areas of interest for GLP-1RA.
Only GP at 24 months and GERD at 12 and 24 months were significantly higher in those on GLP-1RA vs those treated with SGLT-2i. The risk of acute pancreatitis and ileus were lower at all time points in the GLP-1RA treated group. No significant differences in biliary outcomes or intestinal obstruction were observed at any time points.
All-Cause Mortality and MACEs in the Treated and Control Groups
All-cause mortality rates were significantly lower in the GLP-1RA treated group at 3 months (OR 0.65, 95% CI: 0.59–0.72), 6 months (OR 0.74, 95% CI: 0.68–0.79), 12 months (OR 0.81, 95% CI: 0.76–0.85), and 24 months (OR 0.83, 95% CI: 0.80–0.87) when compared with patients on SGLT-2i (Table 4 and Figure 2). Similarly, patients prescribed a GLP-1RA had a reduced risk of MI and stroke at 3, 6, 12, and 24 months (Table 4).
Table 4.
Mortality and MACE Outcomes in T2D on GLP-1RA vs SGLT-2i (After Propensity Score Matching)
| Time since index date (mo) | T2D + GLP-1RA (n) |
T2D + SGLT-2i (n) |
aOR | 95% CI |
|---|---|---|---|---|
| Mortality | ||||
| 3 | 622 | 950 | 0.65 | (0.59–0.72) |
| 6 | 1231 | 1672 | 0.74 | (0.68–0.79) |
| 12 | 2385 | 2945 | 0.81 | (0.76–0.85) |
| 24 | 3998 | 4776 | 0.83 | (0.80–0.87) |
| MI | ||||
| 3 | 451 | 723 | 0.61 | (0.54–0.69) |
| 6 | 774 | 1141 | 0.66 | (0.61–0.73) |
| 12 | 1332 | 1819 | 0.71 | (0.67–0.77) |
| 24 | 2205 | 2719 | 0.79 | (0.75–0.84) |
| Stroke | ||||
| 3 | 470 | 664 | 0.70 | (0.63–0.79) |
| 6 | 793 | 1046 | 0.75 | (0.69–0.83) |
| 12 | 1324 | 1649 | 0.80 | (0.74–0.86) |
| 24 | 2114 | 2405 | 0.87 | (0.82–0.93) |
N = number of subjects with the respective event. 95% CI are provided for each column. All-cause mortality, MI, and Stroke were significantly lower in the GLP-1RA treated group throughout the 2-year period.
Figure 2.
All-cause mortality in T2D on GLP-1RA vs SGLT-2i (after propensity score matching). All-cause mortality was 17% (aOR 0.83; 95% CI 0.79, 0.86) lower in the GLP-1RA treated group compared to SGLT-2i group at 2 years.
Discussion
This real-world study, utilizing a large sample size, found that new diagnoses of GP and GERD were higher in patients prescribed a GLP-1RA compared to those prescribed a SGLT-2i. However, new diagnoses of acute pancreatitis and ileus were lower in the GLP-1RA group while new GI diagnoses such as cholecystitis, cholangitis, or intestinal obstruction were not significantly different. Furthermore, all-cause mortality, diagnosis of acute MI, and stroke were all lowered as early as 3 months of the initiation of GLP treatment and remained lower in the GLP-treated group at 2 years (P < .0001). We showed that all-cause mortality was lower by 17% in patients with T2D treated with GLP-1RA compared to SLGT-2i at 2 years despite having suboptimal glucose control and this is consistent with previous studies.25 Diagnoses of acute MI and stroke were reduced by 21% and 13% at 2 years respectively.
There are limited real-world data on the risk of GI AEs with GLP-1RA in T2D.13,17 We found an increased risk in new diagnosis of GP after 24 months of being prescribed a GLP-1RA and compared to patients prescribed a SGLT-2i. GLP-1RA is associated with delayed gastric emptying in contrast to SGLT-2i which have not been associated with delayed gastric emptying. Clinically we find GP occurs earlier but there may be a delay in diagnosis and gastric emptying testing may not have been performed right away though this specific information is not able to be ascertained from the database. In addition, there may be an effect of change in dosing or higher dosing that delayed gastric emptying. Furthermore, increased risk for new diagnosis of GERD in the GLP-1RA group occurred at 12 and 24 months compared to the SGLT-2i group. This may be partly related to an increased risk of GP. Increased risk of GERD in T2D taking GLP-1RA has been described previously with short-acting GLP-1RA.18 Our study included both short and long-acting GLP-1RA, latter which have been shown to demonstrate milder GI symptoms and greater glucose control.26
Studies assessing the risk of acute pancreatitis in adults with T2D on GLP-1RA have yielded conflicting results.19, 20, 21,27 A recent retrospective chart review of electronic health records at Cleveland Clinic did not show a higher risk of acute pancreatitis in patients with T2D taking GLP-1RA and a previous history of acute pancreatitis.27 Pooled analysis of cardiovascular outcome trials as well as systematic review of meta-analyses did not suggest increased risk of acute pancreatitis with GLP-1RA treatment in patients with T2D.20,21 A population-based cohort study reported an increased risk of hospitalization for acute pancreatitis with older short-acting GLP-1RA exenatide treated patients with T2D;19 however, the number of GLP-1RA users was small limiting statistical power to detect a difference. In contrast, our study, which included both short and long-acting GLP-1RA and larger sample size, did not identify a significant pancreatitis risk.
We also observed an insignificant risk of new diagnosis of biliary outcomes and intestinal obstruction. A nationwide cohort study reported liraglutide but not dulaglutide was associated with an elevated risk of biliary-related diseases compared to SGLT-2i in Asian patients with T2D22 but this risk was different in a diverse population with a heterogeneous array of GLP-1RA and predominantly long-acting GLP-1RA as in our study. Another systematic review and meta-analysis of randomized controlled trials found that use of GLP-1RA was associated with increased risk of gallbladder and biliary diseases but this was when GLP-1RA were used at higher doses and for longer durations for weight loss.23 When pooled with lower doses and shorter duration of GLP-1RA associated with less weight loss we found the risk of biliary disease was insignificant.
Lastly, in a large real-world study, GLP-1RA and dipeptidyl peptidase-4 inhibitors were associated with an increased risk of intestinal obstruction but this association was highest with 1.6–1.8 years of use.24 With shorter duration or lower dose of GLP-1RA, there was no significant increase in obstruction. Mohit et al found that use of GLP-1RA for weight loss (not T2D) compared with use of bupropion-naltrexone was associated with increased risk of pancreatitis, GP, and bowel obstruction but not biliary disease.13 The outcomes for T2D as in our study may be different than for weight loss, which is what the Mohit study evaluated.
The strengths of this study are (1) large database with patients from multiple institutions within the United States. (2) In addition, we analyzed multiple GLP-1RA analogs including semaglutide and tirzepatide over a 2-year period at multiple time intervals. (3) We had a diverse patient population with over 30% under-represented minorities. (4) Propensity matching for demographics, co-morbidities and medications that may have otherwise affected outcomes.
The limitations of this study include: (1) we could not confirm the diagnosis of GP with a gastric emptying study. (2) Patients may have been lost to follow-up due to various reasons, eg transitioning to HCOs outside of the TriNetX network. (3) This is a pooled analysis of all GLP-1RA, but it is possible that different GLP-1RA have different degrees of benefits or AEs. (4) It is not possible to know from the data alone whether patients actually filled the GLP-1RA prescriptions, treatment duration or dose, medication discontinuation, all of which may impact the interpretation of results. (5) We don’t know if the population was privately insured, self-pay or Medicare/Medicaid. (6) Approximately 26% of patients on SGLT-2i were not matched with GLP-1RA patients, which introduces potential selection bias and limits the external validity of the study.
Conclusion
This is a large real-world study of GI AEs and all-cause mortality of GLP-1RA analog treatment for patients with T2D. Patients treated with GLP-1RA had increased risk of GP and GERD as well as decreased risk of MACEs and overall mortality events compared to those treated with SGLT-2i. The diabetes and gastroenterology communities, including regulatory agencies and payers, should consider these findings when assessing efficacy and AEs of these therapeutic agents. Greater awareness by both patients and providers can minimize and treat GI AEs associated with GLP-1RA use. However, future large prospective long-term studies are needed to substantiate our conclusions.
Acknowledgments
The authors extend their thanks to James Bena, MS, Department of Gastroenterology, Hepatology and Nutrition, Digestive Disease Institute, Cleveland Clinic Foundation, Cleveland, OH, USA.
Authors' Contributions
Samita Garg, MD – Conceptualization; Methodology; Supervision; Writing – original draft; Writing – review & editing. Thabet Qapaja – Data curation; Formal analysis; Investigation; Methodology; Writing – review & editing. Osama Hamid – Formal analysis; Figure creation; Writing – review & editing. Gizem Kaya – Writing – review & editing. Rashid Abdel-Razeq – Data tables and analysis. Dina Alayan – Data tables and analysis. Mohammed Abu-Rumaileh – Data table creation and analysis. Anthony Lembo – Writing – review & editing. Steven Nissen – Conceptualization; Investigation; Methodology; Supervision; Writing – review & editing.
Footnotes
Conflicts of Interest: These authors disclose the following: Samita Garg reports receiving an IIS grant from Abbott, a research grant with Enterra, and serving as a speaker/consultant for Novo Nordisk and Ardelyx. Anthony Lembo is a consultant to Takeda, Ironwood, Ardelyx, Vibrant, Evoke, and Atmo, and holds stock in Johnson & Johnson, Allurion, and Bristol Myers Squibb. Steven Nissen reports that the Cleveland Clinic Center for Clinical Research has received funding to conduct clinical trials from AbbVie, AstraZeneca, Amgen, Arrowhead, Bristol Myers Squibb, Eli Lilly, Esperion, Medtronic, MyoKardia, New Amsterdam Pharmaceuticals, Novartis, Pfizer, and Silence Therapeutics. Dr Nissen is involved in these trials but receives no personal remuneration for his participation. He consults for these companies but does not accept compensation. The remaining authors disclose no conflicts.
Funding: The authors report no funding.
Ethical Statement: Institutional review board approval was not required for this study, as it involved analysis of an anonymous database and did not involve any human subject identifiers.
Data Transparency Statement: The data, analytic methods, and study materials used in this research will be made available to other researchers. Resources will be available upon reasonable request.
Reporting Guidelines: This study was a retrospective observational cohort analysis conducted using the TriNetX research database. To ensure transparency, rigor, and reproducibility in reporting, we followed the Strengthening the Reporting of Observational Studies in Epidemiology guidelines, which are specifically designed for observational research. The STROBE checklist was reviewed and applied in accordance with recommendations from the EQUATOR Network (www.equator-network.org) and the Gastro Hep Advances Guide for Authors.
Material associated with this article can be found in the online version at https://doi.org/10.1016/j.gastha.2025.100736.
Supplementary Material
References
- 1.Hong S., Saeedi P., Karuranga S., et al. IDF Diabetes Atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183 doi: 10.1016/j.diabres.2021.109119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention National Diabetes Statistics Report website. https://www.cdc.gov/diabetes/data/statistics-report/index.html
- 3.Statistics about diabetes. ADA. https://diabetes.org/about-diabetes/statistics/about-diabetes
- 4.Ong K.L., Stafford L.K., McLaughlin S.A., et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet. 2023;402(10397):203–234. doi: 10.1016/S0140-6736(23)01301-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.CDC adult obesity facts. https://www.cdc.gov/obesity/data/adult.html/
- 6.Marx N., Husain M., Lehrke M., et al. GLP-1 receptor agonists for the reduction of atherosclerotic cardiovascular risk in patients with type 2 diabetes. Circulation. 2022;146(24):1882–1894. doi: 10.1161/CIRCULATIONAHA.122.059595. [DOI] [PubMed] [Google Scholar]
- 7.Ussher J.R., Drucker D.J. Glucagon-like peptide 1 receptor agonists: cardiovascular benefits and mechanisms of action. Nat Rev Cardiol. 2023;20(7):463–474. doi: 10.1038/s41569-023-00849-3. [DOI] [PubMed] [Google Scholar]
- 8.Marso S.P., Daniels G.H., Brown-Frandsen K., et al. Liraglutide and cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2016;375(4):311–322. doi: 10.1056/NEJMoa1603827. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Marso S.P., Bain S.C., Consoli A., et al. Semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2016;375(19):1834–1844. doi: 10.1056/NEJMoa1607141. [DOI] [PubMed] [Google Scholar]
- 10.Kristensen S.L., Rorth R., Jhund P.S., 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(10):776–785. doi: 10.1016/S2213-8587(19)30249-9. [DOI] [PubMed] [Google Scholar]
- 11.Mansoor M., Birkenfeld A.L., Donsmark M., et al. Oral semaglutide and cardiovascular outcomes in patients with type 2 diabetes. N Engl J Med. 2019;381(9):841–851. doi: 10.1056/NEJMoa1901118. [DOI] [PubMed] [Google Scholar]
- 12.McGuire D.K., Shih W.J., Cosentino F., et al. Association of SGLT2 inhibitors with cardiovascular and kidney outcomes in patients with type 2 diabetes A meta-analysis. JAMA Cardiol. 2021;6(2):148–158. doi: 10.1001/jamacardio.2020.4511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sodhi M., Rezaeianzadeh R., Kezouh A., et al. Risk of gastrointestinal adverse events associated with glucagon-like peptide-1 receptor agonists for weight loss. JAMA. 2023;330(18):1795–1797. doi: 10.1001/jama.2023.19574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Aldhaleei W.A., Abegaz T.M., Bhagavathula A.S. Glucagon-like peptide-1 receptor agonists associated gastrointestinal adverse events: a cross-sectional analysis of the National Institutes of Health all of Us cohort. Pharmaceuticals (Basel) 2024;17(2):199. doi: 10.3390/ph17020199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Osei S.P., Akomaning E., Florut T.F., et al. Gastrointestinal safety assessment of GLP-1 receptor agonists in the US: a real-world adverse events analysis from the FAERS database. Diagnostics (Basel) 2024;14(24):2829. doi: 10.3390/diagnostics14242829. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Liu L., Chen J., Wang L., et al. Association between different GLP-1 receptor agonists and gastrointestinal adverse reactions: a real-world disproportionality study based on FDA adverse event reporting system database. Front Endocrinol (Lausanne) 2022;13 doi: 10.3389/fendo.2022.1043789. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kalas M.A., Galura G.M., McCallum R.W. Medication-induced gastroparesis: a case report. J Investig Med High Impact Case Rep. 2021;9 doi: 10.1177/23247096211051919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Liu B.D., Udemba S.C., Liang K., et al. Shorter-acting glucagon-like peptide-1 receptor agonists are associated with increased development of gastro-oesophageal reflux disease and its complications in patients with type 2 diabetes mellitus: a population-level retrospective matched cohort study. Gut. 2024;73(2):246–254. doi: 10.1136/gutjnl-2023-329651. [DOI] [PubMed] [Google Scholar]
- 19.Singh S., Chang H., Richards T.M., et al. Glucagonlike peptide 1–based therapies and risk of hospitalization for acute pancreatitis in type 2 diabetes mellitus: a population-based matched case-control study. JAMA Intern Med. 2013;173(7):534–539. doi: 10.1001/jamainternmed.2013.2720. [DOI] [PubMed] [Google Scholar]
- 20.Chuqing C., Yang S., Zhou Z. GLP-1 receptor agonists and pancreatic safety concerns in type 2 diabetic patients: data from cardiovascular outcome trials. Endocrine. 2020;68:518–525. doi: 10.1007/s12020-020-02223-6. [DOI] [PubMed] [Google Scholar]
- 21.Li L., Shen J., Bala M.M., et al. Incretin treatment and risk of pancreatitis in patients with type 2 diabetes mellitus: systematic review and meta-analysis of randomised and non-randomised studies. BMJ. 2014;348 doi: 10.1136/bmj.g2366. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dong Y., Wu J., Chang C., et al. Association between glucagon-like peptide-1 receptor agonists and biliary-related diseases in patients with type 2 diabetes: a nationwide cohort study. Pharmacotherapy. 2022;42(6):483–494. doi: 10.1002/phar.2688. [DOI] [PubMed] [Google Scholar]
- 23.He L., Wang J., Ping F., et al. Association of glucagon-like peptide-1 receptor agonist use with risk of gallbladder and biliary diseases: a systematic review and meta-analysis of randomized clinical trials. JAMA Intern Med. 2022;182(5):513–519. doi: 10.1001/jamainternmed.2022.0338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Faillie J., Yin H., Yu O.H.Y., et al. Incretin-based drugs and risk of intestinal obstruction among patients with type 2 diabetes. Clin Pharmacol Ther. 2022;111(1):272–282. doi: 10.1002/cpt.2430. [DOI] [PubMed] [Google Scholar]
- 25.Baviera M., Foresta A., Colacioppo P., et al. Effectiveness and safety of GLP-1 receptor agonists versus SGLT-2 inhibitors in type 2 diabetes: an Italian cohort study. Cardiovasc Diabetol. 2022;21:162. doi: 10.1186/s12933-022-01572-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Garber A.J. Long-acting glucagon-like peptide 1 receptor agonists. A review of their efficacy and tolerability. Diabetes Care. 2011;34(Suppl 2):S279–S284. doi: 10.2337/dc11-s231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Lomeli L.D., Kodali A.M., Tsushima Y., et al. The incidence of acute pancreatitis with GLP-1 receptor agonist therapy in individuals with a known history of pancreatitis. Diabetes Res Clin Pract. 2024;215 doi: 10.1016/j.diabres.2024.111806. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.