By Theo Loxley
Published 27 March 2026  |  TherapyInsights

For more than 60 years, metformin has been one of the most widely prescribed medications in the world a cheap, reliable treatment for type 2 diabetes with an enviable safety record. But a growing body of research is forcing scientists to reconsider what this drug actually does, and who it might help.

In 2025 and into 2026, a series of landmark studies have revealed that metformin acts directly on the brain, may reduce the risk of long COVID, and could slow key markers of biological aging. The findings have placed a decades-old generic medication at the centre of some of the most pressing questions in modern medicine.

None of this means metformin is a miracle drug. But the emerging evidence is significant enough that researchers, clinicians, and policymakers are paying close attention.

Metformin belongs to a class of drugs called biguanides. It was first synthesised from compounds found in the French lilac plant and entered clinical use in France in the late 1950s. By the 1990s, it had become the global first-line treatment for type 2 diabetes, recommended by virtually every major medical guideline.

Its primary job has always been to lower blood sugar. It does this mainly by reducing glucose production in the liver and improving the body’s sensitivity to insulin. It is affordable, widely available, and carries relatively few serious side effects for most patients.

Today, metformin is prescribed to an estimated 150 million people worldwide. It appears on the World Health Organization’s List of Essential Medicines. In Australia, it remains the most commonly dispensed diabetes medication on the Pharmaceutical Benefits Scheme.

What has changed is not the drug itself, it is what scientists are discovering about how it works and what else it may do.

One of the most striking developments came from a 2025 study published in Science Advances by researchers at Baylor College of Medicine in the United States. The team, led by Dr Makoto Fukuda, discovered that metformin acts directly on a specific region of the brain called the ventromedial hypothalamus (VMH), a finding that challenges the long-held assumption that the drug works only in the liver and gut.

The study found that metformin suppresses a protein called Rap1 in the VMH, activating a group of neurons known as SF1 neurons. When researchers removed Rap1 from the brains of genetically modified mice, metformin stopped working, even though other diabetes medications continued to be effective. Remarkably, when tiny doses of metformin were delivered directly into the brain, blood sugar dropped significantly, even at concentrations thousands of times lower than a standard oral dose.

The implications extend beyond diabetes management. The Baylor team has indicated it will now investigate whether this same brain pathway is responsible for metformin’s observed effects on cognitive decline and brain aging.

Separately, a 2026 study published in Communications Medicine used computational modelling to compare 39 diabetes therapies for their potential to protect against Alzheimer’s disease. Metformin emerged as the most promising candidate, outperforming newer drugs including semaglutide. A UK-based cohort study of more than 210,000 participants, published in Alzheimer’s & Dementia in 2025, found that longer duration of metformin use was consistently associated with greater reduction in dementia risk.

An 18-month clinical trial, the Metformin in Alzheimer’s Dementia Prevention (MAP) study is currently underway at Columbia University, with results expected in 2027. It is testing whether metformin can slow cognitive decline in people with mild cognitive impairment who do not have diabetes.

The picture is not straightforward, however. A 2024 study in Translational Psychiatry found that long-term metformin use actually impaired cognition in aged Alzheimer’s model mice, suggesting the drug’s effects may depend on timing, dose, and the presence of existing disease. This kind of complexity is exactly why researchers urge caution in drawing premature conclusions.

Metformin has also attracted significant attention for its potential role in preventing long COVID, the persistent, often debilitating condition that can follow a SARS-CoV-2 infection.

The most influential evidence comes from the COVID-OUT trial, a phase 3 randomised controlled trial conducted across six US sites. Published in The Lancet Infectious Diseases in 2023, it found that a 14-day course of metformin, started within days of a COVID-19 diagnosis, reduced the incidence of long COVID by approximately 41 per cent in adults with overweight or obesity over a 10-month follow-up period.

Since then, the evidence base has strengthened. A large UK retrospective study published in Clinical Infectious Diseases in 2025 analysed data from more than 624,000 patients. It found that early metformin use was associated with a 64 per cent reduction in long COVID risk at one year in adults with overweight or obesity. The ACTIV-6 randomised trial, published in 2025, reported similar findings.

An editorial published in Clinical Infectious Diseases in March 2026 argued that the combined evidence from two large randomised trials and two electronic health record analyses now supports adding metformin to clinical treatment guidelines for non-hospitalised adults with acute COVID-19 infection a step that has not yet been widely adopted.

Data from the NIH’s RECOVER Initiative, the largest US study of long COVID, found that adults already taking metformin for diabetes were less likely to develop long COVID than those taking other diabetes medications.

Scientists believe metformin’s anti-inflammatory properties may play a role. Laboratory studies have shown the drug can restore cellular processes disrupted by SARS-CoV-2 infection and prevent virus-induced cell ageing in neurons. However, researchers emphasise that the strongest evidence applies to specific populations primarily adults with overweight or obesity and further trials are needed to establish whether the benefit extends more broadly.

Perhaps the most ambitious line of inquiry involves metformin’s potential to slow biological aging itself.

A 2024 study published in Cell reported that metformin slowed biological aging clocks in male primates, including a notable regression in brain aging markers. A 2025 study in The Journals of Gerontology, examining data from the Women’s Health Initiative, found that postmenopausal women with type 2 diabetes who took metformin had a 30 per cent lower risk of dying before the age of 90 compared with women taking another diabetes drug, sulfonylurea.

These findings have fuelled interest in the Targeting Aging with Metformin (TAME) trial, a proposed multi-site clinical trial that would enrol 3,000 adults aged 65 to 79 and track whether metformin delays the onset of age-related conditions including cardiovascular disease, cancer, dementia, and death. Designed by researchers at the Albert Einstein College of Medicine, TAME would be the first clinical trial to treat aging itself as a targetable condition.

However, the trial has faced persistent funding challenges. Because metformin is a generic drug with no patent protection, pharmaceutical companies have little financial incentive to sponsor the research. As of late 2025, the trial remained only partially funded, with estimated costs between US$45 million and US$70 million.

Not all scientists are convinced. A review published in Ageing Research Reviews in mid-2025 noted that metformin has generally failed to demonstrate its anticipated benefits in clinical trials involving non-diabetic populations. The authors described an “emerging uncertainty” around metformin’s anti-aging potential and called for more rigorous evidence.

Dr Nir Barzilai, the principal investigator behind TAME and one of the world’s foremost aging researchers, has acknowledged the complexity but maintains that the weight of evidence favours investigation. In an August 2025 interview, he noted that the TAME framework has already influenced the field Eli Lilly is now reportedly designing a similar trial using a GLP-1 agonist, using the regulatory template TAME established with the FDA.

What makes metformin unusual is not any single finding, but the breadth of its apparent effects across very different disease areas.

At the cellular level, metformin activates a key enzyme called AMPK, which plays a central role in how cells manage energy. It also reduces activity along the mTOR pathway, a signalling system involved in cell growth and aging. These twin effects appear to reduce inflammation, limit oxidative stress, and support healthier cellular function processes that are implicated in conditions ranging from diabetes to dementia to cardiovascular disease.

Metformin has also been shown to influence gene activity linked to longevity, limit DNA damage, and reduce markers of cellular senescence the state in which cells stop dividing but continue to produce inflammatory signals.

Put simply, metformin appears to affect several of the fundamental biological processes that underpin chronic disease. That is what makes it so interesting to researchers and so difficult to categorise.

The potential significance of metformin research extends well beyond the laboratory. If even a fraction of the emerging evidence translates into clinical practice, the implications for healthcare systems could be substantial.

Chronic diseases already account for an overwhelming share of healthcare spending in developed nations. In Australia, chronic conditions were responsible for 91 per cent of non-fatal disease burden and 78 per cent of fatal disease burden in 2023, according to the Australian Institute of Health and Welfare. The years lived with disability due to chronic conditions rose from 1.6 million in 2003 to 2.5 million in 2023.

Population aging is compounding these pressures. Australian health expenditure for adults over 65 is projected to nearly double in real terms by 2035. Meanwhile, diabetes prevalence in Australia has increased by 220 per cent over the past two decades, and overweight and obesity surpassed tobacco as the leading risk factor for death in 2024.

A safe, affordable, off-patent medication that could meaningfully reduce the incidence of even one or two major chronic conditions would represent an extraordinary return on investment. Metformin costs a fraction of newer therapies including GLP-1 receptor agonists, which can cost thousands of dollars per year and is already integrated into healthcare systems worldwide. For many patients, the rising cost of accessing healthcare makes affordable interventions especially relevant.

This is why the TAME trial matters beyond its immediate scientific question. If it demonstrates that metformin can delay the onset of multiple age-related diseases simultaneously, it would establish a regulatory and conceptual framework for treating aging as a modifiable condition, rather than an inevitable decline.

For Australia specifically, the metformin research intersects with several major policy challenges.

The National Disability Insurance Scheme (NDIS) supports more than 600,000 participants, many of whom live with conditions that overlap with the areas metformin research is now exploring including cognitive impairment, neurological conditions, and the long-term consequences of chronic disease.

Aged care is under similar strain. Australia’s population of adults aged 85 and over is growing faster than any other age group, driving demand for residential and home-based care services. Conditions like dementia which affects an estimated 400,000 Australians are among the largest cost drivers in the aged care system.

If metformin or similar interventions can delay cognitive decline, reduce the burden of long COVID on people with existing disabilities, or compress the period of age-related morbidity, the downstream effects on both the NDIS and aged care could be meaningful. Fewer years of severe disability, delayed onset of dementia, and reduced post-viral complications all translate to lower long-term support costs and better quality of life.

This is not a reason to overstate the evidence. But it is a reason for Australian policymakers and healthcare planners to monitor the research closely and consider how emerging findings might inform prevention strategies, prescribing guidelines, and system-level planning.

It is important to be clear about what the current evidence does and does not support.

Metformin is approved in Australia and globally for the treatment of type 2 diabetes. It is not currently approved for the prevention of dementia, long COVID, or aging. None of the emerging research has yet changed prescribing guidelines for non-diabetic populations.

The evidence linking metformin to brain health, while promising, is largely based on animal studies, observational data, and computational analyses. The strongest clinical trial data relates to long COVID prevention, but even there, the findings apply primarily to adults with overweight or obesity, and the drug has not yet been incorporated into mainstream COVID-19 treatment guidelines.

Metformin is not without side effects. Gastrointestinal symptoms including nausea, diarrhoea, and abdominal discomfort are common, particularly during the early weeks of use. Long-term use has been associated with vitamin B12 deficiency and, in rare cases, a condition called lactic acidosis, though the risk of the latter is now considered very low in appropriately screened patients.

Metformin is not suitable for everyone. It should be used with caution in patients with significant kidney impairment and is generally not recommended for individuals with advanced liver disease or a history of metabolic acidosis.

Patients should not begin taking metformin for any off-label purpose without consulting their healthcare provider. Self-prescribing based on preliminary research findings carries real risks and may not be appropriate for an individual’s specific health circumstances.

Metformin is being re-examined, not redefined at least, not yet.

The coming years will be critical. The MAP trial results, expected in 2027, will provide the first rigorous clinical evidence on whether metformin can protect against cognitive decline in non-diabetic adults. The fate of the TAME trial still seeking full funding will determine whether the broader question of metformin and aging can be answered definitively. Meanwhile, ongoing analyses of long COVID data continue to build the case for metformin’s role in post-viral illness.

What is already clear is that a drug once viewed as a straightforward glucose-lowering agent is far more complex than anyone expected. The discovery that metformin acts through the brain, combined with accumulating evidence across multiple disease areas, has placed it at the intersection of diabetes research, neuroscience, gerontology, and public health.

Whether metformin ultimately proves to be a transformative multi-disease intervention or a promising lead that falls short in clinical trials, the research has already shifted how scientists think about chronic disease, aging, and the untapped potential of existing medications.

For healthcare systems under growing pressure from aging populations and rising chronic disease burden, that shift in thinking may be just as important as the drug itself.

About the Author

Theo Loxley is a healthcare journalist at TherapyInsights covering NDIS, aged care, and the real-world impact of policy on Australian health services.