Metformin in non-diabetic adults, what the longevity evidence does and does not support

Key takeaways

• Metformin's primary mechanism is mitochondrial complex I inhibition leading to AMPK activation and mTORC1 inhibition; it does not directly raise serum insulin and does not cause hypoglycemia in non-diabetic adults at standard doses.
• The TAME trial (Targeting Aging with Metformin), originally proposed by Nir Barzilai and colleagues at Albert Einstein, was registered as ClinicalTrials.gov NCT04141293 and is intended to test whether metformin delays age-related multi-morbidity. It has not yet completed primary outcome reporting at scale.
• MILES (Metformin in Longevity Study) reported transcriptomic and muscle-metabolism findings in older non-diabetic adults; metformin did not improve insulin sensitivity in this population in some sub-analyses and may modestly attenuate exercise-induced mitochondrial adaptation.
• Observational studies (Bannister et al. 2014, retrospective cohorts) have reported associations between metformin use in type 2 diabetes and lower all-cause mortality and lower cancer incidence vs. other antidiabetics, but the published evidence is observational and confounded by indication.
• Metformin carries a labeled black-box warning on lactic acidosis, a rare but serious adverse event, with risk concentrated in patients with renal impairment, hepatic dysfunction, hypoxia, or sepsis. The label sets specific eGFR cutoffs for use.

Mechanism, in plain terms

Metformin reduces hepatic gluconeogenesis through several pathways. The principal mechanism is partial inhibition of mitochondrial complex I in hepatocytes, which raises the AMP/ATP ratio and activates AMP-activated protein kinase (AMPK). AMPK activation has downstream effects including inhibition of mammalian target of rapamycin complex 1 (mTORC1), suppression of fatty acid synthesis, and improved insulin sensitivity in peripheral tissue. There is also a direct effect on glucagon signaling and a gut-microbiome contribution that has been characterized in more recent work .

Metformin does not stimulate insulin secretion, which is why it does not cause hypoglycemia in non-diabetic adults at standard doses. The clinical practice of starting at 500 mg with food and titrating to 1500 to 2000 mg per day is largely a strategy to manage GI tolerability, which is the most common reason for discontinuation .

UKPDS, the foundational diabetes evidence

The UK Prospective Diabetes Study (UKPDS) randomized newly diagnosed adults with type 2 diabetes to metformin or conventional management. The metformin arm showed a 36 percent reduction in all-cause mortality and a 39 percent reduction in myocardial infarction over a median 10.7 years follow-up. The benefit on cardiovascular endpoints persisted in the 10-year post-trial follow-up .

UKPDS established metformin as first-line therapy for type 2 diabetes, but it does not directly inform metformin use in non-diabetic adults. The cardiovascular benefit observed was in adults with established diabetes and elevated baseline cardiovascular risk.

TAME and other non-diabetic trials

TAME (Targeting Aging with Metformin) is a multicenter, double-blind, randomized trial proposed to enroll approximately 3,000 adults aged 65 to 79 without diabetes, randomized to metformin or placebo for 6 years. The primary endpoint is a composite of major age-related disease incidence (cancer, cardiovascular disease, dementia) plus mortality. The trial is registered on ClinicalTrials.gov; funding and operational status have evolved over time .

MILES (Metformin in Longevity Study) randomized 53 older non-diabetic adults to metformin 1700 mg per day or placebo for 6 weeks. The trial reported transcriptomic effects in skeletal muscle and adipose tissue consistent with metformin's known mechanism. Insulin sensitivity did not improve in non-diabetic adults in this sample size, and effects on mitochondrial biogenesis were modest .

Other named trials in non-diabetic adults with prediabetes (DPP and DPP Outcomes Study) demonstrated diabetes-incidence reduction, but the populations were not aging-cohort populations. The DPP follow-up reported sustained diabetes-incidence reduction over 15 years .

The exercise-adaptation interaction

Metformin's AMPK-activating mechanism overlaps with the AMPK signal that exercise produces in skeletal muscle. Several trials have asked whether metformin co-administered with exercise blunts the cardiometabolic adaptations to training. Konopka et al. 2019 randomized 53 older adults to 12 weeks of supervised aerobic exercise with metformin or placebo. The metformin arm showed attenuated improvements in VO2max and skeletal muscle mitochondrial respiration vs. placebo plus exercise .

Subsequent trials and meta-analyses have produced mixed results, with some trials showing attenuation of training adaptation and others showing minimal interaction. The evidence is consistent with a modest negative effect of metformin on the magnitude of exercise-induced adaptation in skeletal muscle, particularly in older adults, with the clinical significance still being characterized .

Cancer incidence and cognitive function

Observational studies have reported associations between metformin use in adults with type 2 diabetes and reduced cancer incidence (Bannister et al. 2014 reported lower all-cause mortality in metformin-treated diabetic adults vs. matched non-diabetic controls). The association is observational and the effect size in randomized trials specifically powered for cancer outcomes has been smaller and inconsistent .

Cognitive function evidence is mixed. Some observational analyses have suggested lower dementia incidence in metformin-treated adults; others have shown no benefit or possible negative association. There is no trial-grade evidence that metformin prevents Alzheimer disease or other dementia in non-diabetic adults .

Safety, the lactic acidosis warning, and renal cutoffs

Metformin's labeled black-box warning is on metformin-associated lactic acidosis (MALA), a rare (less than 0.1 cases per 1000 patient-years) but serious adverse event. Risk is concentrated in patients with renal impairment, hepatic dysfunction, hypoxia, sepsis, or excessive alcohol use. The label sets eGFR cutoffs: contraindicated below eGFR 30 mL/min/1.73 m^2; not recommended to start below eGFR 45; dose adjustment and increased monitoring at eGFR 30 to 45 .

GI intolerance (diarrhea, abdominal discomfort, nausea) is the most common reason for discontinuation. Slow titration with food and the extended-release formulation reduce GI events. Vitamin B12 deficiency develops in a subset of long-term users (about 5 to 10 percent over years) and is monitorable with serum B12 .

What the evidence does not show

• Metformin has not been shown to extend lifespan or healthspan in non-diabetic adults in any completed trial. TAME is intended to address this; it has not reported primary outcomes.
• Metformin is not approved for any non-diabetic indication. Off-label use in non-diabetic adults is a clinical and prescriber decision.
• Metformin does not consistently improve insulin sensitivity or cardiometabolic markers in metabolically healthy non-diabetic adults at standard doses.
• Metformin may attenuate exercise-induced cardiorespiratory and mitochondrial adaptations in older adults; the trade-off vs. potential AMPK-mediated longevity benefits is not resolved.