GLP-1 Drugs and PAD: Better Walking, Unproven Amputation Prevention

GLP-1 receptor agonists have a credible role in vascular care for people with type 2 diabetes and peripheral artery disease (PAD). The best evidence says they can improve walking performance and reduce broad cardiovascular risk. It does not yet show that they prevent amputations.

The distinction matters because the headline figure is irresistible: a 2026 meta-analysis reported a 47% lower relative risk of lower-extremity amputation among GLP-1RA users. But that estimate came from observational cohorts, not randomized trials, carried very low certainty, and disappeared in a key sensitivity analysis. A large relative-risk reduction can make excellent copy. It cannot substitute for a causal answer.

The most defensible PAD benefit is walking

In a 2026 systematic review and meta-analysis by Giugliano et al., two randomized controlled trials included 847 people with type 2 diabetes and PAD who received liraglutide or semaglutide. Across those trials, GLP-1RA treatment was associated with improved functional walking distance: hazard ratio (HR) 1.10, 95% confidence interval (CI) 1.05 to 1.15; p<0.001. Statistical heterogeneity was zero (I²=0.0%), and the authors rated the certainty of this evidence moderate.

That is a real controlled signal, but it needs plain limits. The analysis reports the result as a relative hazard ratio for improvement, not as additional meters on a treadmill or six-minute walk test. Readers therefore cannot compare it directly with supervised exercise therapy or revascularization. The result supports better walking function; it does not tell us how much farther a typical patient walked.

The clinical setting also matters. This evidence concerns people with both type 2 diabetes and established PAD, not healthy people taking semaglutide for weight loss and hoping for quieter arteries. No randomized trial in asymptomatic or subclinical atherosclerosis shows that GLP-1 drugs prevent vascular disease.

Why the amputation number does not settle the question

The same Giugliano review pooled four cohort studies involving 104,292 patients and found lower-extremity amputation was less frequent among GLP-1RA users (risk ratio 0.53, 95% CI 0.39 to 0.73; p<0.001). That is the source of the 47% relative reduction.

Its weaknesses are unusually consequential. The studies were highly inconsistent (I²=85.96%), with one study driving much of the result, and the GRADE assessment rated the amputation evidence very low certainty. More tellingly, the apparent benefit disappeared when analyses excluded patients with diabetic foot ulcers at baseline.

That pattern fits prescribing-channeling bias. Clinicians may reasonably avoid or defer a GLP-1RA in patients with an active foot ulcer, who also face the highest near-term amputation risk. If treated patients begin healthier than untreated patients in ways a database cannot fully measure, the drug can look protective even when it is not the cause of the difference. Observational adjustment helps, but it cannot reliably erase this kind of clinical selection.

GLP-1RAs are associated with fewer amputations in some cohorts. They are not proven to prevent amputations.

Randomized data sharpen the problem. In two placebo-controlled trials of exenatide and semaglutide, totaling 3,592 participants, GLP-1RAs did not significantly reduce major adverse limb events or lower-extremity amputations. These were not dedicated severe-PAD trials and were not powered specifically for limb outcomes, so the result does not prove the drugs lack a limb benefit. It does mean the controlled evidence has not demonstrated one.

The cohort literature remains useful context, not a verdict. A 2025 Journal of Vascular Surgery cohort study by Go et al. found GLP-1RA use was associated with fewer major adverse cardiovascular and limb events in people with moderate PAD. Hsiao et al. in JAMA Network Open likewise reported an association between GLP-1RA use and fewer subsequent major adverse limb events among people with diabetes who had a prior limb event. These findings point in the same direction, but they share the fundamental limitation of nonrandomized treatment assignment.

An active-comparator analysis offers a little more perspective. Lin et al. found fewer cardiovascular and limb events among GLP-1RA users than among users of DPP-4 inhibitors in a 2022 diabetes cohort. In a separate comparison with SGLT2 inhibitors, GLP-1RA outcomes were comparable rather than clearly better. Both comparisons remain observational, but the latter result is a useful warning against treating every GLP-1RA-versus-no-treatment contrast as a drug-specific limb effect.

What guidelines actually support

The 2024 ACC/AHA PAD guidance recommends GLP-1RAs for people with type 2 diabetes and PAD to reduce major adverse cardiovascular events, as discussed in the Hsiao et al. analysis. ADA recommendations for people with diabetes and established atherosclerotic cardiovascular disease also support use of cardioprotective agents, including GLP-1RAs; a 2025 post-bypass cohort study describes that context.

Those recommendations do not create an amputation-prevention indication. They reflect the established cardiovascular benefits of this drug class in the appropriate high-risk population. A GLP-1RA belongs beside smoking cessation, lipid lowering, blood-pressure treatment, glucose management, foot surveillance, exercise when feasible, and revascularization assessment when indicated. It does not replace any of them.

HFpEF: meaningful symptoms, unfinished outcome questions

The story is more encouraging, though still incomplete, in obesity-related heart failure with preserved ejection fraction (HFpEF). A 2026 systematic review and meta-analysis by Sebastian et al. concluded that tirzepatide improved efficacy and safety outcomes in this population. The available publication does not fully resolve whether reported event outcomes were hard endpoints such as cardiovascular death and hospitalization or composites that included health-status and functional measures. That distinction matters: a higher symptom score or longer walk is valuable, but it does not automatically predict fewer deaths.

A 2026 federated real-world study by Monzo et al. compared semaglutide with tirzepatide in people with obesity and HFpEF. Both drugs were associated with weight loss, improved health status, and lower clinical event rates. Tirzepatide did not show statistically significant superiority over semaglutide on clinical event rates.

Tirzepatide's dual GIP/GLP-1 mechanism is biologically interesting, but mechanism is not an outcome trial. At the time these publications were compiled, SURPASS-CVOT, the dedicated cardiovascular outcomes trial comparing tirzepatide with dulaglutide, had not reported primary results. Claims that tirzepatide is the superior vascular drug should wait for evidence designed to answer that question.

The practical verdict

For a person with type 2 diabetes and diagnosed PAD, GLP-1RAs have guideline support for cardiovascular risk reduction and moderate-certainty randomized evidence for improved walking performance. Discussing one with a clinician can be reasonable within comprehensive vascular care.

For obesity-related HFpEF, semaglutide and tirzepatide appear to improve weight and health status, with emerging evidence of fewer clinical events. Specialist supervision remains appropriate because comparative hard-endpoint evidence is unsettled.

For everyone else, do not turn a PAD cohort finding into a preventive vascular prescription. The credible claim is narrower and still useful: these drugs can improve metabolic health, reduce broad cardiovascular risk in indicated patients, and may help walking in diabetes with PAD. Whether they prevent amputations remains an unanswered trial question.