Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Dec 4.
Published in final edited form as: Diabetes Res Clin Pract. 2025 Nov 1;230:112982. doi: 10.1016/j.diabres.2025.112982

Risk of Lower Extremity Complications with GLP-1 Receptor Agonists, SGLT2 Inhibitors, and DPP-4 Inhibitors in Peripheral Artery Disease

Alexander T Hong 1, Forest Lin 2, Ivan Y Luu 1, Laura Shin 1, Sukgu Han 1, David G Armstrong 1, Tze-Woei Tan 1
PMCID: PMC12645714  NIHMSID: NIHMS2121410  PMID: 41177308

Abstract

Aims:

To compare the association of glucagon-like peptide-1 receptor agonists (GLP-1RA), sodium-glucose cotransporter-2 inhibitors (SGLT2i), and dipeptidyl peptidase-4 inhibitors (DPP-4i) therapies on lower extremity vascular complications and mortality in PAD.

Methods:

We conducted a retrospective cohort study using a nationwide U.S. electronic health records database. Adults with PAD and type 2 diabetes who initiated GLP-1RA, SGLT2i, or DPP4i therapy (May 2013-January 2025) were included. Propensity score matching (1:1) balanced baseline characteristics. Cox models estimated hazard ratios (HR) with 95% confidence intervals (CI) for major amputation (primary outcome), lower extremity revascularization (LER), and all-cause mortality over 3 years follow-up. Subgroup analyses included symptomatic PAD, prior LER, and drug-level comparisons.

Results:

GLP-1RAs were associated with lower risks of major amputation (HR 0.79 [95% CI 0.70–0.90]), LER (HR 0.82 [0.76–0.88]), and mortality (HR 0.71 [0.68–0.73]) compared with SGLT2i (n=77,393 each). Similar reductions were seen versus DPP-4is (n=39,907 each); SGLT2is and DPP-4is showed comparable risks (n=42,924 each). GLP-1RA benefits remained significant in symptomatic PAD (p<0.05) and were associated with lower mortality in LER patients (p<0.05). Semaglutide and tirzepatide showed the greatest benefit.

Conclusion:

GLP-1RAs were associated with lower limb complication and mortality risks in PAD, supporting a potential vascular benefit that warrants prospective evaluation.

Keywords: diabetes mellitus, dipeptidyl peptidase-4 inhibitors, glucagon-like peptide-1 receptor agonists, limb amputation, peripheral artery disease, sodium-glucose cotransporter-2 inhibitors

1.1.1. INTRODUCTION

Lower extremity complications are among the most debilitating consequences of type 2 diabetes mellitus, with over four-fold increased risk of lower extremity amputation (LEA) and more than double the rate of lower extremity revascularization (LER) among those with concomitant peripheral artery disease (PAD).1–7 Globally, PAD affects more than 113 million people, with 70% whom have disease attributable to modifiable risk factors including type 2 diabetes.8,9 PAD significantly increases mortality risk in type 2 diabetes, ranging from 10% increase in in-hospital mortality to a 2-fold higher risk of long-term all-cause mortality.2,10–13 Beyond its clinical toll, PAD management also imposes significant economic burden, with annual U.S. healthcare costs exceeding $84 billion.14,15 As the global prevalence of PAD projected to double by 2050, there is an urgent need to identify glucose-lowering therapies that offer both effective metabolic control and vascular protection.8,16,17

Among modern antihyperglycemic agents, glucagon-like peptide-1 receptor agonists (GLP-1RA) and sodium-glucose cotransporter-2 inhibitors (SGLT2i) have demonstrated cardiovascular and renal benefits beyond glycemic control.18,19 GLP-1RAs also exhibit anti-inflammatory and endothelial-stabilizing effects, supporting interest in their potential for limb protection.20,21 In contrast, SGLT2is have been scrutinized following early reports of increased amputation risk.22,23 Despite growing evidence for cardiovascular outcomes, data on comparing the effects of these agents on lower extremity complications in PAD remain limited. Dipeptidyl peptidase-4 inhibitors (DPP-4i), which increase endogenous GLP-1 by preventing its degradation, have shown some benefit on limb complications in type 2 diabetes but have not been specifically studied in PAD.24–26 Patients with symptomatic PAD, including those with rest pain, tissue loss, or prior LER, face substantial risk for adverse limb events yet remain understudied. It is unknown whether GLP-1RA, SGLT2i, or DPP-4i differentially affect the risks of amputation, gangrene, LER, and mortality, leaving a critical gap in guiding therapy selection for PAD patients. This study leverages a large, real-world dataset to compare lower extremity complication risks across these drug classes, with focused analyses in high-risk PAD subgroups.

1.2.1. METHODS

1.2.2. Data Source

This retrospective cohort study utilized the U.S. Collaborative Network within the TriNetX Analytics platform, a federated health research network aggregating de-identified electronic health record (EHR) data from more than 124 million patients across 72 US healthcare organizations, including academic and non-academic institutions.27 The study received exemption from the University of Southern California Institutional Review Board HS-24–00606, and informed consent was waived due to the use of de-identified data. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology reporting guidelines.28

1.2.3. Study Population

Cohort selection is detailed in Figure 1. Individuals aged 18 years or older with type 2 diabetes and PAD were identified using International Statistical Classification of Diseases, Tenth Revision (ICD-10) codes. Validated codes from prior studies was adapted to identify PAD, LER, and LEA.21,29,30 Analysis was performed on June 9th 2025, covering a study period from May 1, 2013, to January 1, 2025, as the first SGLT2i received FDA-approved in March 2013. The index event was defined as the first prescription of GLP-1RA, SGLT2i, or DPP-4i therapy. A new-user, active comparator design was applied to reduce immortal time bias, restricting inclusion to patients with no prior exposure to any of the study drug classes. Eligible patients were categorized into three mutually exclusive cohorts: (1) GLP-1RA only, (2) SGLT2i only, or (3) DPP-4i only. Patients could switch medications within the same drug class (e.g., between different GLP-1RAs, SGLT2is, or DPP-4is). To avoid confounding from other antihyperglycemic agents, patients were required to initiate GLP-1RA, SGLT2i, or DPP-4i therapy without new prescriptions for other oral glucose-lowering medications (biguanides, alpha-glucosidase inhibitors, thiazolidinediones, or sulfonylureas) within three months prior to or three years following the index date. Additionally, patients in each cohort could not have prior exposure to, or initiate within three years after the index date, agents from the other two study drug classes (e.g., GLP-1RA users could not use SGLT2is or DPP-4is). Participants were followed from the date of initial prescription until the earliest occurrence of the following: 1) outcome event, 2) death, 3) loss to follow-up, or 4) three years post-index date. A list of codes used in analysis can be found in Supplementary Table 1.

Figure 1. Study cohort flow diagram.

Figure 1.

Open in a new tab

PAD: Peripheral artery disease. GLP-1RA: Glucagon-like peptide-1 receptor agonists. SGLT2i: Sodium-glucose cotransporter-2 inhibitors. DPP-4i: Dipeptidyl dipeptidase-4 inhibitor.

1.2.4. Propensity Score Matching

Propensity score matching was used to adjust for confounding by balancing clinical variables and risk factors for LEA and LER. Three matched cohorts were created to assess the effects of GLP-1RAs relative to SGLT2is and DPP-4is: 1) GLP-1RA only versus (vs.) SGLT2i only; 2) SGLT2i only vs. DPP-4i only; and 3) GLP-1RA only vs. DPP-4i only. Covariates for matching included demographics (age at index, sex, race, ethnicity); comorbidities (cardiovascular diseases, chronic kidney disease, dementia, chronic lower respiratory disease, diabetic angiopathy, diabetic neuropathy, liver cirrhosis, neoplasms, socioeconomic and psychosocial health hazards, nicotine/tobacco use); medications (insulin, other hypoglycemic agents, lipid-lowering agents, antihypertensives, antithrombotic agents, antiarrhythmic agents); laboratory values (body mass index [BMI], glycosylated hemoglobin, cholesterol, glomerular filtration rate [GFR], triglycerides); and healthcare utilization (emergency department, inpatient, and outpatient visits). Diagnoses used for propensity score matching, including codes referencing gangrene, were restricted to those recorded prior to or on the index date and were not included in outcome analyses unless they occurred during follow-up. Detailed codes are provided in Supplementary Table 2.

1.2.5. Outcomes

The primary outcome was major LEA (i.e., above-ankle)-free survival. Secondary outcomes included any (i.e., above or below-ankle) LEA, lower extremity gangrene, LER, and all-cause mortality. Outcomes were evaluated at 3 years following the index prescription. Outcome definitions were based on code lists validated in prior studies.21,31

1.2.6. Statistical Analysis

All statistical analyses were performed using the analytical tools available within the TriNetX platform. Continuous variables were reported as means with standard deviations (SD), while categorical variables were presented as counts and percentages. To inform variable selection and ensure alignment with clinical reasoning, a conceptual framework was developed to map hypothesized relationships among exposures, comorbidities, and lower extremity outcomes (Supplemental Figure 1). Propensity scores were generated via logistic regression using prespecified covariates. A 1:1 greedy nearest-neighbor matching approach was applied, with a caliper width of 0.1 pooled SDs of the logit of the propensity score. Matching was performed using available patient characteristics. Missing data were not imputed, and variables with incomplete data could not be directly used for matching unless the data were present. Cohort balance post-matching was evaluated using standardized mean differences (SMD), with values below 0.1 considered adequate balance. Kaplan-Meier survival curves were used to estimate survival probabilities, and cohort differences were assessed using log-rank tests. Number-at-risk counts were derived based on the number of patients with follow-up at each time interval who were event-free. Cox proportional hazards analysis was conducted to calculate hazard ratios (HRs) with 95% confidence intervals (CIs) for each outcome. Proportional hazards assumptions were evaluated via scaled Schoenfeld residuals in R (version 3.2–3). Statistical significance was defined as a two-sided P-value <0.05.

Three prespecified subgroup analyses were performed to enhance the robustness of the findings: 1) those with symptomatic PAD (defined as claudication, rest pain, ulceration, or gangrene) at treatment initiation, excluding gangrene as an outcome, 2) those with history of LER procedures, excluding LER as an outcome, and 3) comparisons of individual agents within each drug class to all other agents in the same class, limited to agents with sufficiently large cohorts to reliably estimate outcomes (i.e., semaglutide and tirzepatide).

1.3.1. RESULTS

1.3.2. Associations between GLP-1RA or SGLT2i and Lower Extremity Outcomes

The GLP-1RA cohort included 99,056 patients (mean [SD] age 64.8 [10.8] years; 47.0% female, 67.2% White, 17.3% Black, and 8.7% Hispanic), and the SGLT2i cohort included 116,756 patients (mean [SD] age 69.1 [10.6] years; 36.4% female, 65.5% White, 17.9% Black, and 8.5% Hispanic). After propensity score matching, 77,393 patients were included in each group with well-balanced covariates (Supplementary Table 3). The median (interquartile range [IQR]) follow-up duration was 1.7 (1.3) years for the GLP-1RA cohort and 1.5 (1.5) years for the SGLT2i cohort.

At 3 years, the GLP-1RA cohort had 435 major amputations (99.06% survival) compared with 521 in the SGLT2i cohort (98.85% survival; P=0.002). (Figure 2). GLP-1RA use was associated with lower risk for major amputation (HR 0.79 [95% CI 0.70, 0.90]) compared with SGLT2i (Figure 3, Supplementary Table 4). Risks were also lower for any amputation (HR 0.86 [95% CI 0.79–0.94]), gangrene (HR 0.81 [95% CI 0.75–0.86]), LER (HR 0.82 [95% CI 0.76, 0.88]), and all-cause mortality (HR 0.71 [95% CI 0.68–0.73]) in the GLP-RA cohort.

Figure 2. Major Amputation-Free Survival at 3 years.

Figure 2.

Open in a new tab

Kaplan-Meier survival analysis of major amputation-free survival among PAD patients receiving GLP-1RA vs SGLT2i (2a) in a 3-year follow up. Separate analyses were conducted patients with GLP-1RA vs DPP-4i (2b) or SGLT2i vs DPP-4i (2c).

PAD: Peripheral artery disease. GLP-1RA: Glucagon-like peptide-1 receptor agonists. SGLT2i: Sodium-glucose cotransporter-2 inhibitors. DPP-4i: Dipeptidyl dipeptidase-4 inhibitor.

Figure 3. Forest Plot of Amputation and Mortality at 3 years.

Figure 3.

Open in a new tab

Patients were followed up for as long as 3 years after the index event for both groups. Hazard ratios (HRs) with 95% confidence intervals (CI) were calculated using a Cox proportional hazards model with censoring applied. For each outcome, the groups were propensity-score matched for covariates related to the outcome, and the outcome was compared between the matched groups. Each eligible individual was followed up from the index event until the occurrence of the outcomes, death, loss to follow-up, or 3 years after the index event, whichever occurred first.

GLP-1RA: Glucagon-like peptide-1 receptor agonists. SGLT2i: Sodium-glucose cotransporter-2 inhibitors. DPP-4i: Dipeptidyl dipeptidase-4 inhibitor.

1.3.3. Associations between GLP-1RA or DPP-4i and Lower Extremity Outcomes

The DPP-4i only cohort included 45,333 patients (mean [SD] age 68.2 [10.6] years; 44.0% female, 60.0% White, 19.2% Black, and 12.1% Hispanic). After propensity score matching, 39,907 patients were included in each group, with all covariates balanced except for GFR (GLP-1RA: 63.1±28.2 vs. DPP-4i: 59.6±31.1, SMD=0.116) (Supplementary Table 5). The DPP-4i cohort had a median (IQR) follow-up of 2.0 (1.0) years.

At 3 years, the GLP-1RA cohort had 259 major amputations (99.01% survival) compared with 378 in the DPP-4i cohort (98.70% survival) (P=0.001) (Figure 2). GLP-RA use was associated with lower risk for major amputation (HR 0.77 [95% CI 0.66–0.90]) compared with DPP-4i (Figure 3, Supplementary Table 4). Risks were also lower for any amputation (HR 0.88 [95% CI 0.78–0.98]), gangrene (HR 0.78 [95% CI 0.71–0.86]), LER (HR 0.80 [95% CI 0.73–0.88]), and mortality (HR 0.57 [95% CI 0.55–0.60]) in the GLP-1RA cohort.

1.3.4. Associations between SGLT2i or DPP-4i and Lower Extremity Outcomes

The SGLT2i cohort included 116,756 patients (mean [SD] age 69.1 [10.6] years; 36.4% female, 65.5% White, 17.9% Black, and 8.5% Hispanic) and the DPP-4i cohort included 45,333 patients (mean [SD] age 68.2 [10.6] years; 44.0% female, 60.0% White, 19.2% Black, and 12.1% Hispanic). The propensity score-matched groups consisted of 42,924 patients in each cohort, with all covariates well balanced except for GFR (SGLT2i: 64.1±27.8 vs. DPP-4i: 59.5±31.0, SMD=0.156) (Supplementary Table 6).

At 3 years, the SGLT2i group had 318 major amputations (98.76% survival) compared with 397 in the DPP-4i cohort (98.73% survival) (P=0.17) (Figure 2). SGLT2i therapy was associated with lower risk of all-cause mortality (HR 0.77 [95% CI 0.74–0.80]), with no significant differences observed for other outcomes (Figure 3, Supplementary Table 4).

1.3.5. Subgroup Analyses

Among individuals with symptomatic PAD, GLP-1RA use was associated with lower risks for major amputation (HR 0.82 [95% CI 0.68–0.99]), LER (HR 0.84 [95% CI 0.74–0.97]), and mortality (HR 0.76 [95% CI 0.69–0.83]) compared with SGLT2i (Table 1). Compared with DPP-4is therapy, GLP-1RA therapy was associated with lower risks of any amputation (HR 0.83 [95% CI 0.68–0.99]), LER (HR 0.79 [95% CI 0.67–0.93]), and mortality (HR 0.63 [95% CI 0.57–0.70]). When comparing SGLT2i with DPP-4i, the only significant association observed was a lower risk of mortality in SGLT2i therapy (HR 0.83 [95% CI 0.76–0.91]).

Table 1.

Lower Extremity and Mortality Outcomes by Treatment Group in Patients with Symptomatic PAD

Treatment Groups Outcome GLP-1RA, n (%) SGLT2i, n (%) HR (95% CI)
GLP-1RA vs. SGLT2i
(n=9086)		 Major LEA 203 (2.4) 231 (2.7) 0.82 (0.68–0.99)
Any LEA 306 (4.0) 335 (4.3) 0.87 (0.75–1.02)
LER 397 (6.7) 432 (7.3) 0.84 (0.74–0.97)
Mortality 878 (9.7) 1052 (12.0) 0.76 (0.69–0.83)
GLP-1RA, n (%) DPP4i, n (%) HR (95% CI)
GLP-1RA vs. DPP-4i
(n=5374)		 Major LEA 134 (2.6) 172 (3.4) 0.84 (0.67–1.05)
Any LEA 194 (4.3) 256 (5.6) 0.83 (0.68–0.99)
LER 246 (7.0) 337 (9.3) 0.79 (0.67–0.93)
Mortality 556 (10.7) 969 (18.6) 0.63 (0.57–0.70)
SGLT2i, n (%) DPP4i, n (%) HR (95% CI)
SGLT2i vs. DPP-4i
(n=5923)		 Major LEA 173 (3.1) 188 (3.4) 1.07 (0.87–1.31)
Any LEA 253 (5.0) 288 (5.7) 1.00 (0.85–1.19)
LER 326 (8.4) 384 (9.7) 0.99 (0.85–1.15)
Mortality 750 (13.1) 1105 (19.2) 0.83 (0.76–0.91)
Open in a new tab

PAD, peripheral artery disease; GLP-1RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose cotransporter 2 inhibitor; DPP-4i, dipeptidyl dipeptidase-4 inhibitor; LEA, lower extremity amputation; LER, lower extremity revascularization; CI, confidence interval; HR, hazard ratio.

Among patients with history of LER, GLP-1RA use was associated with lower risks of gangrene (HR 0.73 [95% CI 0.59–0.90]) and mortality (HR 0.68 [95% CI 0.60–0.77]) compared with the SGLT2i cohort (Table 2). Compared with DPP-4is, GLP-1RA use was associated with lower risks for major amputation (HR 0.70 [95% CI 0.53–0.94]) and mortality (HR 0.65 [95% CI 0.56–0.76]). Risk of lower extremity outcomes and mortality were similar between matched SGLT2i and DPP4i cohorts among patients with history of LER.

Table 2.

Lower Extremity Outcomes and Mortality by Treatment Group in Patients with PAD and History of Lower Extremity Revascularization

Treatment Groups Outcome GLP-1RA, n (%) SGLT2i, n (%) HR (95% CI)
GLP-1RA vs. SGLT2i
(n=4241)		 Major LEA 127 (3.3) 135 (3.5) 0.88 (0.69–1.12)
Any LEA 179 (5.5) 182 (5.5) 0.96 (0.78–1.18)
Gangrene 153 (5.8) 200 (7.6) 0.73 (0.59–0.90)
Mortality 447 (10.9) 597 (14.4) 0.68 (0.60–0.77)
GLP-1RA, n (%) DPP4i, n (%) HR (95% CI)
GLP-1RA vs. DPP-4i
(n=2265)		 Major LEA 76 (3.7) 115 (5.6) 0.70 (0.53–0.94)
Any LEA 109 (6.2) 126 (7.3) 0.92 (0.71–1.19)
Gangrene 87 (6.2) 118 (8.8) 0.80 (0.60–1.05)
Mortality 265 (12.1) 441 (20.0) 0.65 (0.56–0.76)
SGLT2i, n (%) DPP4i, n (%) HR (95% CI)
SGLT2i vs. DPP-4i
(n=2556)		 Major LEA 98 (4.2) 128 (5.5) 0.87 (0.67–1.13)
Any LEA 114 (5.8) 144 (7.3) 0.88 (0.69–1.12)
Gangrene 119 (7.9) 131 (8.8) 1.07 (0.84–1.38)
Mortality 378 (15.1) 507 (20.4) 0.89 (0.78–1.01)
Open in a new tab

PAD, peripheral artery disease; GLP-1RA, glucagon-like peptide-1 receptor agonist; SGLT2i, sodium-glucose cotransporter 2 inhibitor; DPP-4i, dipeptidyl dipeptidase-4 inhibitor; LEA, lower extremity amputation; CI, confidence interval; HR, hazard ratio.

When comparing individual agents within drug classes, tirzepatide use was associated with lower risks of any amputation (HR 0.77 [95% CI 0.63–0.96]), LER (HR 0.73 [95% CI 0.60–0.89]), and mortality (HR 0.50 [95% CI 0.45–0.56]) compared to other GLP-1 RA agents (Supplementary Table 7). Semaglutide was similarly associated with reduced risks of major amputation (HR 0.76 [95% CI 0.63–0.91]), gangrene (HR 0.86 [95% CI 0.78–0.95]), LER (HR 0.85 [95% CI 0.76–0.94]), and mortality (HR 0.77 [95% CI 0.73–0.81]). In contrast, both dulaglutide (HR 1.31 [95% CI 1.23–1.39]) and liraglutide (HR 1.33 [95% CI 1.20–1.47]) were associated with increased mortality risk.

Among SGLT-2i agents, empagliflozin use was associated with lower risks of gangrene (HR 0.88 [95% CI 0.80–0.96]), LER (HR 0.88 [95% CI 0.80–0.98]), and mortality (HR 0.90 [95% CI 0.86–0.93]) compared to other SGLT-2is (Supplementary Table 8). Lower extremity outcome risks were comparable for both canagliflozin and dapagliflozin.

Among DPP-4is, sitagliptin use was associated with lower mortality (HR 0.85 [95% CI 0.81–0.90]) compared to other DPP-4is, while other lower extremity complication risks were comparable between sitagliptin and linagliptin (Supplementary Table 9).

1.4.1. DISCUSSION

In this large, real-world cohort of patients with type 2 diabetes and PAD, GLP-1RA therapy was associated with an estimated 20% lower risk of lower extremity amputation, gangrene, LER, and mortality compared with SGLT2i and DPP-4i. These benefits remained evident in patients with symptomatic PAD. In individuals with prior LER, GLP-1RAs were associated with lower risks of mortality and gangrene, though results were less consistent across endpoints and comparators. Among individual agents, tirzepatide and semaglutide were associated with greatest reductions in lower extremity complications among GLP-1RAs. Collectively, the results highlight the potential role for GLP-1RA in limb preservation among high-risk patients with type 2 diabetes and PAD.

Despite their growing adoption as second-line therapies in T2D, evidence on the comparative effects of GLP-1RA, SGLT2i, and DPP-4i therapies in patients with established PAD remains extremely limited. While numerous studies have examined PAD as an outcome, few have specifically stratified by PAD status—a key distinction given the heightened risk of limb complications in this subgroup. Existing randomized controlled trial (RCT) data on SGLT2 inhibitors have not demonstrated a consistent increase in limb-related adverse events in PAD. Empagliflozin was associated with a non-significant reduction in amputation risk (HR 0.84 [95% CI 0.54–1.32]); canagliflozin showed comparable amputation rates versus placebo (23.96 vs. 20.90 events per 1,000 patient-years; HR 1.09 [95% CI 0.82–1.46]; P = 0.864); and dapagliflozin demonstrated a non-significant increase in amputation risk (HR 1.51 [95% CI 0.94–2.42]) with no differences in LER (HR 0.86 [95% CI 0.65–1.15]).16,32–34 Similarly, exenatide showed no difference in LEA rates compared to placebo, occurring in 5.0% vs. 4.9% of patients, respectively (HR 0.99 [95% CI 0.71–1.38]).35 To our knowledge, no other RCTs have evaluated the effects of other GLP-1RAs or DPP-4is on lower extremity outcomes specifically in PAD.

Our study found that among GLP-1 RAs, tirzepatide and semaglutide were associated with significantly lower risks of amputation, LER, and mortality, including a 24% reduction in major amputation risk with semaglutide and a 23% reduction in any amputation with tirzepatide. Dulaglutide and liraglutide, though similar in risk of amputation, were both associated with over 30% increased mortality relative to their class. Empagliflozin was associated with lower risks of gangrene, LER, and mortality compared to other SGLT-2is, supporting RCT findings of a potential amputation risk reduction. In contrast, canagliflozin and dapagliflozin conferred no observable limb benefit, reinforcing their neutral safety profiles. However, the relatively small sample size of canagliflozin users in our cohort may have limited the power to detect modest differences in outcomes. Among DPP-4is, sitagliptin and linagliptin showed no significant differences in lower extremity outcomes. These findings highlight the clinical relevance of agent-level selection within drug classes and underscore the need for PAD-specific considerations in glucose-lowering therapy to optimize limb preservation and survival.

Our prior study was among the first to show that GLP-1RA therapy was associated with a lower risk of amputation compared to SGLT2is in patients with type 2 diabetes.21 Similarly, a recent study of moderate PAD found reduced major adverse limb events with GLP-1RA use versus non-users, but lacked comparisons with other second-line therapies.36 However, neither study assessed high-risk PAD subgroups (e.g., symptomatic PAD or prior LER), compared GLP-1RA to DPP-4 inhibitors — a class still frequently used in clinical practice — or elucidated drug-specific differences within classes. The present analysis extends prior evidence by demonstrating that GLP-1RA therapy is associated with significantly lower risks of amputation and mortality compared to DPP-4is in patients with PAD, with consistent benefits observed in high-risk PAD subgroups. These findings are consistent with prior studies in broader type 2 diabetes cohorts.25,26 We also observed lower risks of gangrene and LER with GLP-1RA therapy, further supporting its potential limb-preserving effects in PAD. In contrast, comparisons between SGLT2i and DPP-4i in prior type 2 diabetes studies have yielded mixed results, with some reporting reduced limb event risk with SGLT2i,37 while others found no significant difference.38 In this PAD-specific analysis, limb outcomes were comparable between SGLT2i and DDP-4i cohorts; however, SGLT-2i therapy was associated with lower mortality, suggesting a potential survival advantage in this high-risk population.

Subgroup analyses further demonstrate the potential benefit of GLP-1RA therapy in more advanced PAD. Among patients with symptomatic PAD, GLP-1RAs were associated with lower risks of major amputation, gangrene, LER, and mortality compared with both SGLT2is and DPP-4is. Although absolute differences in amputation-free survival at three years were modest, the consistency across limb and survival outcomes supports a clinically meaningful effect. In patients with history of LER, GLP-1RA therapy remained associated with lower risks of major amputation and mortality compared with DPP-4i, and lower risks of gangrene and mortality compared with SGLT2i. No significant differences in limb outcomes were observed between SGLT2i and DPP-4i cohorts in either subgroup. These findings suggest that GLP-1RA therapy may offer incremental benefits in limb preservation in patients with symptomatic or surgically treated PAD, an important benefit in populations with limited effective medical options. Prospective trials in PAD-specific populations are needed to validate these associations and to define the therapeutic role of GLP-1RAs in reducing limb complications.

Several potential mechanisms may underlie the association between GLP-1RA therapy and reduced risk of LEAs and related vascular outcomes, independent of glycemic control. These agents enhance endothelial function by promoting nitric oxide production and reducing oxidative stress, while suppressing pro-inflammatory cytokines such as TNF-α and IL-1β, affecting key mediators of atherosclerosis and PAD progression.39,40 They may also inhibit vascular smooth muscle proliferation, reduce matrix metalloproteinase activity, and stabilize atherosclerotic plaques.41 Additionally, GLP-1RAs have been shown to promote angiogenesis and improve microvascular perfusion, both of which are critical in mitigating ischemic injury and promoting wound healing in PAD.42 Indirect benefits such as weight loss and improved insulin sensitivity may further contribute to slowing the progression of both microvascular and macrovascular complications.43

Several limitations warrant consideration. First, as an observational study, causal inference cannot be established, and residual confounding may persist despite extensive propensity score matching. Key variables such as baseline PAD severity, ulcer characteristics (size, depth, or infection status), wound management, and revascularization techniques were not available and may have influenced outcomes or treatment selection, introducing residual confounding. Without access to these granular clinical details, observed differences in outcomes may partly reflect underlying disease severity rather than the direct effects of the medications. Diabetes duration, glycemic variability, and other metabolic factors were also not measured and may contribute to confounding. Second, outcome miscalculation due to coding inaccuracies and underreporting of lower extremity events are possible. Third, medication adherence, switching, discontinuation, or treatment duration could not be reliably assessed. Fourth, due to analytic platform constraints, we were unable to perform a competing risks analysis treating mortality as a competing event; however, patients who died during follow-up were censored at the time of death in our survival models. Fifth, follow-up duration was limited due to the relatively recent adoption of GLP-1RAs and SGLT2is in clinical practice, restricting our ability to assess long-term outcomes. Finally, although the study leveraged real-world cohorts, these findings may not be generalizable to populations outside the United States or to healthcare systems with different practice patterns, medication accessibility, or PAD treatment protocols. Future prospective studies with longer follow-up, detailed clinical phenotypes, medication adherence tracking, and granular data on diabetes severity may help address these limitations and better define the mechanisms underlying the potential limb-protective effects of GLP-1RAs.

In this national cohort of adults with type 2 diabetes and PAD, treatment with GLP-1RA was associated with lower risks of limb events and all-cause mortality compared with SGLT2is and DPP-4is, with the greatest benefits observed for semaglutide and tirzepatide. These associations remained consistent in patients with symptomatic PAD and prior revascularization. Given the substantial burden of limb-related morbidity and mortality in PAD, selecting therapies with potential vascular benefit may be needed to optimize outcomes. Prospective trials targeting PAD populations and incorporating limb-specific endpoints are needed to establish the role of GLP-1RAs in reducing adverse limb outcomes in this high-risk population.

Supplementary Material

1

HIGHLIGHTS.

  • GLP-1RA therapy linked to lower amputation risk in PAD versus SGLT2i/DPP-4i

  • Lower risks of gangrene, revascularization, and mortality seen with GLP-1RA therapy

  • Semaglutide and tirzepatide showed strongest protective associations among GLP-1RAs

  • Benefits of GLP-1RA consistent among symptomatic PAD and prior revascularization

  • SGLT2i and DPP-4i showed comparable risks for lower limb outcomes

ACKNOWLEDGEMENTS

Funding Source:

This study is partially supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases Award Number 1K23DK122126 and 1R03DK140420.

Roles of the Funders:

The sponsor or funding organization had no role in the design or conduct of this research.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Declarations of Interest: none

Meeting Presentation: This work has been accepted for full oral presentation at Clinical Congress 2025, scheduled for October 4–7th, 2025 in Chicago, IL.

REFERENCES

  • 1.Barnes JA, Eid MA, Creager MA, Goodney PP. Epidemiology and Risk of Amputation in Patients With Diabetes Mellitus and Peripheral Artery Disease. Arterioscler Thromb Vasc Biol. 2020;40(8):1808–1817. doi: 10.1161/ATVBAHA.120.314595 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Criqui MH, Matsushita K, Aboyans V, et al. Lower Extremity Peripheral Artery Disease: Contemporary Epidemiology, Management Gaps, and Future Directions: A Scientific Statement From the American Heart Association. Circulation. 2021;144(9):e171–e191. doi: 10.1161/CIR.0000000000001005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Weissler EH, Clare RM, Lokhnygina Y, et al. Predicting Major Adverse Limb Events in Individuals With Type 2 Diabetes: Insights From the EXSCEL Trial. Diabet Med. 2021;38(10):e14552. doi: 10.1111/dme.14552 [DOI] [Google Scholar]
  • 4.Shammas AN, Jeon-Slaughter H, Tsai S, et al. Major Limb Outcomes Following Lower Extremity Endovascular Revascularization in Patients With and Without Diabetes Mellitus. J Endovasc Ther. 2017;24(3):376–382. doi: 10.1177/1526602817705135 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Fadini GP, Spinetti G, Santopaolo M, Madeddu P. Impaired Regeneration Contributes to Poor Outcomes in Diabetic Peripheral Artery Disease. Arterioscler Thromb Vasc Biol. 2020;40(1):34–44. doi: 10.1161/ATVBAHA.119.312863 [DOI] [PubMed] [Google Scholar]
  • 6.Gyldenkerne C, Olesen KKW, Thrane PG, et al. Trends in Peripheral Artery Disease, Lower-Extremity Revascularization, and Lower-Extremity Amputation in Incident Type 2 Diabetes: A Danish Population-Based Cohort Study. Diabetes Care. 2025;48(1):76–83. doi: 10.2337/dc24-1644 [DOI] [PubMed] [Google Scholar]
  • 7.Jude EB, Oyibo SO, Chalmers N, Boulton AJ. Peripheral arterial disease in diabetic and nondiabetic patients: a comparison of severity and outcome. Diabetes Care. 2001;24(8):1433–1437. doi: 10.2337/diacare.24.8.1433 [DOI] [PubMed] [Google Scholar]
  • 8.Deng L, Du C, Liu L, et al. Forecasting the Global Burden of Peripheral Artery Disease from 2021 to 2050: A Population-Based Study. Research (Wash D C). 8:0702. doi: 10.34133/research.0702 [DOI] [Google Scholar]
  • 9.GBD 2019 Peripheral Artery Disease Collaborators. Global burden of peripheral artery disease and its risk factors, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet Glob Health. 2023;11(10):e1553–e1565. doi: 10.1016/S2214-109X(23)00355-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Keller K, Schmitt VH, Vosseler M, et al. Diabetes Mellitus and Its Impact on Patient-Profile and In-Hospital Outcomes in Peripheral Artery Disease. J Clin Med. 2021;10(21):5033. doi: 10.3390/jcm10215033 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Avdic T, Carlsen HK, Rawshani A, Gudbjörnsdottir S, Mandalenakis Z, Eliasson B. Risk factors for and risk of all-cause and atherosclerotic cardiovascular disease mortality in people with type 2 diabetes and peripheral artery disease: an observational, register-based cohort study. Cardiovasc Diabetol. 2024;23(1):127. doi: 10.1186/s12933-024-02226-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Vrsalovic M, Vucur K, Vrsalovic Presecki A, Fabijanic D, Milosevic M. Impact of diabetes on mortality in peripheral artery disease: a meta-analysis. Clin Cardiol. 2017;40(5):287–291. doi: 10.1002/clc.22657 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Mueller T, Hinterreiter F, Poelz W, Haltmayer M, Dieplinger B. Mortality rates at 10 years are higher in diabetic than in non-diabetic patients with chronic lower extremity peripheral arterial disease. Vasc Med. 2016;21(5):445–452. doi: 10.1177/1358863X16643603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hirsch AT, Hartman L, Town RJ, Virnig BA. National health care costs of peripheral arterial disease in the Medicare population. Vasc Med. 2008;13(3):209–215. doi: 10.1177/1358863X08089277 [DOI] [PubMed] [Google Scholar]
  • 15.McDermott KM, Bose S, Keegan A, Hicks CW. Disparities in Limb Preservation and Associated Socioeconomic Burden among Patients with Diabetes and/or Peripheral Artery Disease in the United States. Semin Vasc Surg. 2023;36(1):39–48. doi: 10.1053/j.semvascsurg.2023.01.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Mahé G, Aboyans V, Cosson E, et al. Challenges and opportunities in the management of type 2 diabetes in patients with lower extremity peripheral artery disease: a tailored diagnosis and treatment review. Cardiovasc Diabetol. 2024;23(1):220. doi: 10.1186/s12933-024-02325-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Aday AW, Matsushita K. Epidemiology of Peripheral Artery Disease and Polyvascular Disease. Circulation Research. 2021;128(12):1818–1832. doi: 10.1161/CIRCRESAHA.121.318535 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Palmer SC, Tendal B, Mustafa RA, et al. Sodium-glucose cotransporter protein-2 (SGLT-2) inhibitors and glucagon-like peptide-1 (GLP-1) receptor agonists for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials. BMJ. 2021;372:m4573. doi: 10.1136/bmj.m4573 [DOI] [Google Scholar]
  • 19.Neuen BL, Fletcher RA, Heath L, et al. Cardiovascular, Kidney, and Safety Outcomes With GLP-1 Receptor Agonists Alone and in Combination With SGLT2 Inhibitors in Type 2 Diabetes: A Systematic Review and Meta-Analysis. Circulation. 2024;150(22):1781–1790. doi: 10.1161/CIRCULATIONAHA.124.071689 [DOI] [PubMed] [Google Scholar]
  • 20.Alharbi SH. Anti-inflammatory role of glucagon-like peptide 1 receptor agonists and its clinical implications. Ther Adv Endocrinol Metab. 2024;15:20420188231222367. doi: 10.1177/20420188231222367 [DOI] [Google Scholar]
  • 21.Hong AT, Luu IY, Lin F, et al. Differential Effect of GLP-1 Receptor Agonists and SGLT2 Inhibitors on Lower-Extremity Amputation Outcomes in Type 2 Diabetes: A Nationwide Retrospective Cohort Study. Diabetes Care. Published online June 25, 2025:dc250292. doi: 10.2337/dc25-0292 [DOI] [Google Scholar]
  • 22.Ueda P, Svanström H, Melbye M, et al. Sodium glucose cotransporter 2 inhibitors and risk of serious adverse events: nationwide register based cohort study. BMJ. 2018;363:k4365. doi: 10.1136/bmj.k4365 [DOI] [Google Scholar]
  • 23.Chang HY, Singh S, Mansour O, Baksh S, Alexander GC. Association Between Sodium-Glucose Cotransporter 2 Inhibitors and Lower Extremity Amputation Among Patients With Type 2 Diabetes. JAMA Intern Med. 2018;178(9):1190–1198. doi: 10.1001/jamainternmed.2018.3034 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.D’Andrea E, Wexler DJ, Kim SC, Paik JM, Alt E, Patorno E. Comparing Effectiveness and Safety of SGLT2 Inhibitors vs DPP-4 Inhibitors in Patients With Type 2 Diabetes and Varying Baseline HbA1c Levels. JAMA Internal Medicine. 2023;183(3):242–254. doi: 10.1001/jamainternmed.2022.6664 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Chang CC, Chen YT, Hsu CY, et al. Dipeptidyl Peptidase-4 Inhibitors, Peripheral Arterial Disease, and Lower Extremity Amputation Risk in Diabetic Patients. Am J Med. 2017;130(3):348–355. doi: 10.1016/j.amjmed.2016.10.016 [DOI] [PubMed] [Google Scholar]
  • 26.Lin DSH, Lee JK, Chen WJ. Major adverse cardiovascular and limb events in patients with diabetes treated with GLP-1 receptor agonists vs DPP-4 inhibitors. Diabetologia. 2021;64(9):1949–1962. doi: 10.1007/s00125-021-05497-1 [DOI] [PubMed] [Google Scholar]
  • 27.Palchuk MB, London JW, Perez-Rey D, et al. A global federated real-world data and analytics platform for research. JAMIA Open. 2023;6(2):ooad035. doi: 10.1093/jamiaopen/ooad035 [DOI] [Google Scholar]
  • 28.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ. 2007;335(7624):806–808. doi: 10.1136/bmj.39335.541782.AD [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Tan TW, Caldwell B, Zhang Y, Kshirsagar O, Cotter DJ, Brewer TW. Foot and Ankle Care by Podiatrists and Amputations in Patients With Diabetes and Kidney Failure. JAMA Netw Open. 2024;7(3):e240801. doi: 10.1001/jamanetworkopen.2024.0801 [DOI] [Google Scholar]
  • 30.Weissler EH, Lippmann SJ, Smerek MM, et al. Model-Based Algorithms for Detecting Peripheral Artery Disease Using Administrative Data From an Electronic Health Record Data System: Algorithm Development Study. JMIR Med Inform. 2020;8(8):e18542. doi: 10.2196/18542 [DOI] [Google Scholar]
  • 31.Cronenwett JL, Birkmeyer JD. Table 6. CPT Codes. October 2000. Accessed October 2, 2025. https://www.ncbi.nlm.nih.gov/books/NBK588524/table/ta6/
  • 32.Bonaca MP, Wiviott SD, Zelniker TA, et al. Dapagliflozin and Cardiac, Kidney, and Limb Outcomes in Patients With and Without Peripheral Artery Disease in DECLARE-TIMI 58. Circulation. 2020;142(8):734–747. doi: 10.1161/CIRCULATIONAHA.119.044775 [DOI] [PubMed] [Google Scholar]
  • 33.Barraclough JY, Yu J, Figtree GA, et al. Cardiovascular and renal outcomes with canagliflozin in patients with peripheral arterial disease: Data from the CANVAS Program and CREDENCE trial. Diabetes Obes Metab. 2022;24(6):1072–1083. doi: 10.1111/dom.14671 [DOI] [PubMed] [Google Scholar]
  • 34.Verma S, Mazer CD, Al-Omran M, et al. Cardiovascular Outcomes and Safety of Empagliflozin in Patients With Type 2 Diabetes Mellitus and Peripheral Artery Disease. Circulation. 2018;137(4):405–407. doi: 10.1161/CIRCULATIONAHA.117.032031 [DOI] [PubMed] [Google Scholar]
  • 35.Badjatiya A, Merrill P, Buse JB, et al. Clinical Outcomes in Patients With Type 2 Diabetes Mellitus and Peripheral Artery Disease: Results From the EXSCEL Trial. Circ Cardiovasc Interv. 2019;12(12):e008018. doi: 10.1161/CIRCINTERVENTIONS.119.008018 [DOI] [Google Scholar]
  • 36.Go CC, Annie F, Drabish K, Eslami MH. Glucagon-like peptide-1 receptor agonists are associated with fewer major adverse cardiovascular and limb events in patients with moderate peripheral arterial disease. J Vasc Surg. Published online June 6, 2025:S0741–5214(25)01112–7. doi: 10.1016/j.jvs.2025.05.037 [DOI] [Google Scholar]
  • 37.Lee HF, Chen SW, Liu JR, et al. Major adverse cardiovascular and limb events in patients with diabetes and concomitant peripheral artery disease treated with sodium glucose cotransporter 2 inhibitor versus dipeptidyl peptidase-4 inhibitor. Cardiovasc Diabetol. 2020;19(1):160. doi: 10.1186/s12933-020-01118-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Lee HF, Chuang C, Li PR, Yeh YH, Chan YH, See LC. Adverse cardiovascular, limb, and renal outcomes in patients with diabetes after peripheral artery disease revascularization treated with sodium glucose cotransporter 2 inhibitors versus dipeptidyl peptidase-4 inhibitors. Diabetology & Metabolic Syndrome. 2023;15(1):8. doi: 10.1186/s13098-023-00982-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Hachuła M, Kosowski M, Ryl S, Basiak M, Okopień B. Impact of Glucagon-Like Peptide 1 Receptor Agonists on Biochemical Markers of the Initiation of Atherosclerotic Process. Int J Mol Sci. 2024;25(3):1854. doi: 10.3390/ijms25031854 [DOI] [Google Scholar]
  • 40.Wu Q, Li D, Huang C, et al. Glucose control independent mechanisms involved in the cardiovascular benefits of glucagon-like peptide-1 receptor agonists. Biomed Pharmacother. 2022;153:113517. doi: 10.1016/j.biopha.2022.113517 [DOI] [Google Scholar]
  • 41.Fan SH, Xiong QF, Wang L, Zhang LH, Shi YW. Glucagon-like peptide 1 treatment reverses vascular remodelling by downregulating matrix metalloproteinase 1 expression through inhibition of the ERK1/2/NF-κB signalling pathway. Mol Cell Endocrinol. 2020;518:111005. doi: 10.1016/j.mce.2020.111005 [DOI] [Google Scholar]
  • 42.Zhang Y, Wang S, Zhou Q, et al. Novel Angiogenesis Role of GLP-1(32–36) to Rescue Diabetic Ischemic Lower Limbs via GLP-1R-Dependent Glycolysis in Mice. Arterioscler Thromb Vasc Biol. 2024;44(6):1225–1245. doi: 10.1161/ATVBAHA.124.320714 [DOI] [PubMed] [Google Scholar]
  • 43.Jiang Y, Wang Z, Ma B, et al. GLP-1 Improves Adipocyte Insulin Sensitivity Following Induction of Endoplasmic Reticulum Stress. Front Pharmacol. 2018;9. doi: 10.3389/fphar.2018.01168 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS2121410-supplement-1.docx (281.1KB, docx)

RESOURCES