Metformin Fails to Improve Walking in PAD Patients Without Diabetes

Peripheral artery disease (PAD) is a common and disabling cardiovascular condition affecting millions of adults, characterized by atherosclerotic obstruction of lower-extremity arteries that leads to impaired walking ability and reduced quality of life. Despite its prevalence, few pharmacological therapies have proven effective at improving functional walking performance in PAD patients. Metformin, a widely available and inexpensive biguanide medication, had emerged as a candidate because of its pleiotropic biological effects — including activation of AMP-activated protein kinase (AMPK), reduction of oxidative stress, and stimulation of endothelial nitric oxide synthase (eNOS) — mechanisms that theoretically could improve skeletal muscle metabolism and vascular function in ischemic limbs.

The PERMET trial was a multicenter, double-blind, placebo-controlled randomized clinical trial conducted across 4 US centers. Enrollment ran from May 2017 to February 2025, with 202 of a targeted 212 participants enrolled (95%), limited by funding constraints. Eligible participants were adults aged 50 and older with confirmed PAD and no diabetes. They were randomized to metformin (n = 97) or matched placebo (n = 105) for 6 months. The mean age was 69.6 years, 28% were female, and 39% were Black. Follow-up was completed August 2025, with 89% (179 of 202) finishing the 6-month assessment.

The primary outcome — change in 6-minute walk distance from baseline to 6 months — showed no significant difference between groups. Both arms declined modestly: metformin group from 358.6 m to 353.2 m (−5.4 m) and placebo group from 359.8 m to 354.5 m (−5.3 m), yielding an adjusted between-group difference of 1.1 m (95% CI, −16.3 to 18.6 m; P = .90). This was far below the minimum clinically important difference of 8 to 20 m. All pre-specified secondary outcomes — maximal and pain-free treadmill walking time, Walking Impairment Questionnaire distance and speed scores, SF-36 physical functioning, and brachial artery flow-mediated dilation — also showed no significant benefit from metformin.

Adverse events were generally consistent with metformin's known profile. Serious adverse events were uncommon and similar between groups (cardiovascular events: 3.1% metformin vs. 1.9% placebo). The most notable nonserious adverse event difference was gastrointestinal: indigestion or stomach upset occurred in 64.9% of metformin participants versus 40.6% of placebo participants. Headache was slightly more common in the placebo group (49.5% vs. 37.2%).

These results have meaningful clinical and research implications. Despite the biological rationale for metformin's potential benefit in PAD — particularly its AMPK-activating and anti-oxidative effects — the trial found no functional benefit over 6 months in a non-diabetic PAD population. The trial was slightly underpowered due to early closure (202 vs. 212 target), but the near-zero between-group difference makes a meaningful benefit unlikely. Researchers and clinicians should look to other therapeutic strategies, such as supervised exercise, gene therapy approaches, or other pharmacological agents, for improving walking performance in PAD.