Metformin and Inflammation: How Does It Affect Your Body?

Metformin, the most widely prescribed drug for type 2 diabetes, doubles as a broad anti-inflammatory agent that damps down harmful immune signaling in blood vessels, fat tissue, the brain, the gut lining, and elsewhere. Its effects on inflammation operate partly through the same energy-sensing pathway it uses to lower blood sugar, and partly through mechanisms researchers are still mapping out. What makes metformin unusual is how many different tissues seem to benefit, and the growing evidence that these anti-inflammatory effects occur even in people who do not have diabetes.

How Metformin Quiets Inflammation at the Cellular Level

Most of metformin’s anti-inflammatory action traces back to a cellular energy sensor called AMPK. When metformin activates AMPK in the lining of blood vessels, it blocks a key inflammatory switch called NF-κB, which normally turns on genes that produce inflammatory proteins and help immune cells stick to vessel walls. In lab experiments on human endothelial cells, metformin shut down NF-κB activation triggered by the inflammatory signal TNF-alpha in a dose-dependent fashion, and when researchers knocked out the AMPK gene, metformin’s ability to suppress that inflammatory response dropped sharply.¹

Metformin also works at a deeper level inside cells by interfering with mitochondria, the structures that generate energy. It partially blocks a component of the mitochondrial machinery called complex I, which reduces the production of reactive oxygen species, the aggressive molecules that fan the flames of inflammation. In immune cells called macrophages, this complex I inhibition lowered production of the potent inflammatory molecule IL-1β while boosting the anti-inflammatory molecule IL-10.² A separate study on intact mitochondria confirmed that metformin’s effect on reactive oxygen species depends on which fuel the mitochondria are burning, inhibiting certain pathways while leaving others unaffected.³

One of the most studied downstream targets is the NLRP3 inflammasome, a multi-protein alarm system inside immune cells that, when activated, triggers the release of IL-1β and can drive tissue damage during severe infections or chronic disease. Metformin suppresses NLRP3 inflammasome activation in lung macrophages, which reduced both IL-1β and IL-6 secretion in models of acute respiratory distress, including one triggered by SARS-CoV-2.⁴ In diabetic heart tissue, the same NLRP3-suppressing effect was confirmed, and it was reversed when an AMPK inhibitor was added, linking the effect directly to metformin’s AMPK activation.⁵

What Happens to Inflammatory Blood Markers

The cellular findings sound promising, but the question most people care about is whether metformin actually lowers measurable inflammation in living humans. The short answer is that it reliably lowers C-reactive protein (CRP), a widely used blood marker of systemic inflammation, but its effect on other markers is less consistent.

A meta-analysis of randomized controlled trials in people with type 2 diabetes found that CRP dropped significantly with metformin treatment, but levels of TNF-α and IL-6 did not change in a statistically meaningful way.⁶ A separate meta-analysis in women with polycystic ovary syndrome (PCOS) found a similar pattern: CRP came down, but IL-6 did not show a reliable change overall, though subgroup analysis suggested IL-6 reductions were more likely in women with higher body mass.⁷

This gap between what happens in a petri dish and what shows up in a blood draw is worth understanding. CRP is made by the liver in response to many upstream inflammatory signals, so it captures a broader picture than any single cytokine. TNF-α and IL-6 circulate at much lower concentrations, fluctuate throughout the day, and can be harder to measure reliably across studies. It does not mean metformin has no effect on those cytokines in specific tissues; it means the effect may not always be large enough or consistent enough to show up in pooled blood samples across diverse patient groups.

Macrophage Polarization and Immune Cell Balance

Macrophages are immune cells that can exist in different functional states. Broadly, M1-type macrophages drive inflammation and fight infections, while M2-type macrophages calm inflammation and promote tissue repair. Metformin consistently nudges macrophages toward the M2 state in chronic inflammatory conditions. In obese mice, metformin lowered blood levels of the inflammatory cytokines IL-6 and TNF-α and reduced M1 macrophage markers in fat tissue, while promoting M2 polarization partly through AMPK activation.⁸ In wound healing models, the same M2 shift accelerated tissue repair and blood vessel formation.⁹

But the dose matters. Research on human neonatal macrophages found that low-dose metformin promoted the anti-inflammatory M2 state, while high doses actually pushed macrophages toward the pro-inflammatory M1 state and worsened inflammation.¹⁰ This dose-dependent flip is an important nuance that gets lost in most popular coverage of metformin as a universal anti-inflammatory. A broader review of the evidence noted something equally striking: in the context of tumors, metformin actually increases M1 macrophages while decreasing M2 macrophages, essentially switching from its usual anti-inflammatory mode to a pro-inflammatory, anti-tumor mode.¹¹ This context-dependent behavior suggests metformin is less of a blunt anti-inflammatory hammer and more of an immune modulator that responds to the surrounding tissue environment.

Metformin also affects T cells, the other major arm of the immune system. It suppresses the proliferation of T cells in general, and specifically dials down Th1 and Th17 cells (types that promote autoimmune and inflammatory responses) while encouraging the development of regulatory T cells (Tregs), which keep immune responses in check.¹² This Th17/Treg rebalancing has been confirmed in liver injury models, where metformin improved outcomes by shifting the balance toward regulatory T cells and reducing inflammatory cytokine release.¹³

The Gut Connection

Metformin’s relationship with the gut is a two-way street. It changes the composition of gut bacteria in ways that may reinforce its anti-inflammatory effects, but those same gut changes also explain the gastrointestinal side effects that bother many users.

People with diabetes who take metformin have higher levels of Akkermansia muciniphila, a bacterium that maintains the gut’s protective mucus layer, along with several bacteria that produce short-chain fatty acids, molecules with their own well-documented anti-inflammatory properties.¹⁴ In older obese mice fed a high-fat diet, metformin increased the number of goblet cells (the cells that produce mucus in the intestinal lining), which strengthened the gut barrier, reduced “leaky gut,” and lowered systemic inflammation. Those mice also showed improved cognitive function, suggesting the gut-barrier repair had effects well beyond the intestines.¹⁵

Interestingly, the gut microbiome shifts may also mediate some of metformin’s anti-inflammatory effects on the brain. In mice with brain hemorrhage, metformin reduced neuroinflammation by promoting the anti-inflammatory state in brain immune cells, and this effect was linked to changes in the gut microbiome, not just to metformin acting directly on brain tissue.¹⁶

Blood Vessels and Heart Health

Chronic inflammation in blood vessel walls is a central driver of atherosclerosis, and metformin appears to intervene at several points along that pathway. In the cells that line blood vessels, metformin reduces oxidative stress and restores the availability of nitric oxide, the molecule that keeps vessels relaxed and flexible. In obese diabetic mice, metformin restored endothelial function through AMPK activation while reducing markers of cellular stress.¹⁷ A focused review of the cardiovascular evidence concluded that metformin improves endothelial dysfunction through multiple targets, both AMPK-dependent and independent, all converging on reduced vascular inflammation and improved vessel health.¹⁸

In diabetic mice prone to atherosclerosis, metformin reduced mitochondrial fragmentation in blood vessel cells, lowered the release of damaging superoxide molecules, improved blood vessel relaxation, and suppressed the growth of atherosclerotic plaques.¹⁹ This is one area where metformin’s anti-inflammatory and metabolic effects are hard to tease apart. Lower blood sugar itself reduces vascular inflammation, so some of the cardiovascular benefit in people with diabetes comes from glucose control rather than a direct anti-inflammatory effect. But as discussed below, evidence from people without diabetes suggests the anti-inflammatory contribution is real and independent.

Effects in the Brain

Neuroinflammation, the activation of the brain’s resident immune cells (microglia), plays a role in stroke, traumatic brain injury, and neurodegenerative conditions. Metformin reduces microglial activation and lowers the production of inflammatory cytokines in brain tissue.²⁰ In a mouse model of neonatal stroke, metformin treatment after the injury did not change the total number of microglia but significantly reduced their activation, which correlated with better behavioral outcomes in the injured mice.²¹

Most of this evidence comes from animal models, so it is too early to say whether metformin would meaningfully slow neuroinflammation in humans with conditions like Alzheimer’s or Parkinson’s disease. But the consistency of the animal findings across different types of brain injury has generated enough interest that clinical trials are underway.

Fat Tissue, Obesity, and Metabolic Inflammation

Excess body fat, especially visceral fat around the organs, is not an inert energy store. It is an active source of inflammatory signals that contribute to insulin resistance, fatty liver disease, and cardiovascular risk. Metformin influences fat tissue through multiple routes: it modulates how fat cells develop, how they release hormones (adipokines), and how immune cells behave within fat depots.²²

In obese mice, metformin reduced macrophage infiltration and inflammatory signaling in brown fat and restored the tissue’s ability to respond to cold exposure, a marker of healthy metabolic function.²³ This matters because when brown fat becomes inflamed, it loses its calorie-burning capacity, contributing to a cycle of worsening obesity and worsening inflammation.

Does It Work If You Do Not Have Diabetes?

This is one of the most-asked questions in the metformin conversation, and the evidence is encouraging. A study that directly tested whether metformin’s anti-inflammatory effects depend on diabetes status concluded that they do not. The anti-inflammatory properties were present regardless of whether someone had diabetes, a finding that has fueled interest in using metformin for non-diabetic cardiovascular disease.²⁴

This independence from blood sugar control aligns with the cellular findings. The AMPK pathway that metformin activates is a general energy-sensing system present in nearly every cell type, not something unique to glucose metabolism. When metformin blocks NF-κB or suppresses the NLRP3 inflammasome, it is acting on inflammatory machinery that operates in anyone’s body, not just in someone with elevated blood sugar. That said, metformin is not approved for anti-inflammatory use in non-diabetic populations, and the long-term risk-benefit balance outside of diabetes and pre-diabetes is still being studied.

Reproductive Health and PCOS

Metformin has been prescribed for PCOS for years, primarily to address insulin resistance. But the anti-inflammatory dimension adds another layer. In PCOS patients, metformin lowered expression of NLRP3 and IL-1β genes and upregulated a micro-RNA (miR-223) known to suppress inflammasome activity, with the effects being most pronounced in obese patients.²⁵ A pilot study also found that metformin reduced serum levels of fractalkine (a chemokine involved in chronic inflammation) and high-sensitivity CRP in PCOS patients, alongside improvements in clinical features like acne, excess hair growth, and insulin resistance.²⁶

Infectious Disease and the Cytokine Storm Question

During the COVID-19 pandemic, metformin attracted attention because severe COVID outcomes often involved an overblown immune response, the so-called cytokine storm. Metformin’s known ability to suppress excessive cytokine production, inhibit neutrophil extracellular traps (the sticky webs that neutrophils deploy during infection but that can clog blood vessels when overproduced), and calm Th1 and Th17 cells all suggested it might help prevent immune overreaction in severe respiratory infections.²⁷ Lab research confirmed that metformin suppressed the inflammatory response of human immune cells exposed to the SARS-CoV-2 spike protein, reducing IL-6 gene expression and dampening both glycolytic and mitochondrial metabolic shifts that fuel the overreaction.²⁸

These are mechanistic and in-vitro findings, and they do not prove that taking metformin will protect you from severe COVID or any other specific infection. But they illustrate a broader principle: metformin’s anti-inflammatory actions could be relevant well beyond metabolic disease, wherever excessive immune activation is part of the problem.

Joints, Cartilage, and Wound Healing

Metformin’s reach extends to tissues that might seem unrelated to a diabetes drug. In joint cartilage, cellular and animal studies show that metformin protects cartilage cells through AMPK activation, suppression of NF-κB, inhibition of enzymes that degrade cartilage, reduction of oxidative stress and inflammation, and shifting of local macrophages toward the anti-inflammatory state.²⁹ Whether this translates to meaningful joint-pain relief in humans with osteoarthritis remains an open question, but the preclinical data are consistent enough to have prompted clinical investigation.

In wound healing, topical metformin applied to diabetic wounds in animal models increased cell proliferation, boosted collagen production, and accelerated closure, essentially reversing the impaired healing that diabetes causes.³⁰ A systematic review of the topical metformin literature confirmed these findings across many studies, noting pro-angiogenic, anti-inflammatory, and antibacterial effects as the main drivers of improved healing.³¹

Aging, Cellular Senescence, and the Longevity Angle

Cellular senescence, the state in which cells stop dividing and start secreting inflammatory molecules, accumulates with age and drives chronic low-grade inflammation sometimes called “inflammaging.” Metformin has attracted interest in the longevity research community partly because it appears to counteract this process. In stem cells from adipose tissue of older donors, metformin reversed oxidative stress, rescued declining cell proliferation, and reduced the accumulation of senescent cells and their associated inflammatory markers. The same treatment had no such effect in cells from younger donors, whose cells were not experiencing the oxidative stress that metformin corrects.³²

This finding is a useful corrective to the hype. Metformin does not appear to be a youth serum that benefits cells regardless of their state. Instead, it rescues cells that are under oxidative stress and heading toward dysfunction, which becomes more common with age but is not universal. The TAME (Targeting Aging with Metformin) trial is underway to test whether metformin can delay age-related diseases in older adults without diabetes, and inflammation is one of the key biomarkers being tracked.

Practical Caveats and Side Effects

Metformin is generally well tolerated, but it is not side-effect-free, and some of its drawbacks are relevant to anyone considering it for its anti-inflammatory potential.

The most common complaints are gastrointestinal: nausea, bloating, diarrhea, and stomach cramps, especially during the first few weeks. Extended-release formulations reduce these issues for many people. The gut microbiome changes that support metformin’s anti-inflammatory effects may partly explain the GI symptoms, as shifts in bacterial populations can produce gas and loose stools before the gut adjusts.

A more insidious concern is vitamin B12 deficiency. Long-term metformin use is associated with impaired B12 absorption, and there is a statistically significant link between metformin use and low B12 levels.³³ One cross-sectional study found that the average duration of metformin use before B12 deficiency developed was roughly 14 years, and the deficiency was associated with peripheral neuropathy, nerve damage that can cause tingling and numbness in the hands and feet.³⁴ Ironically, B12 deficiency itself can drive inflammation and neurological symptoms, so periodic B12 monitoring is recommended for anyone on metformin long-term.

Metformin’s Dual Role in Cancer Immunology

The tumor microenvironment presents an unusual test case for metformin’s immune effects. Tumors often co-opt M2 macrophages and regulatory T cells to create a locally immunosuppressive environment that shields the tumor from attack. In this setting, metformin’s typical anti-inflammatory tendencies could theoretically be harmful. But evidence suggests the opposite happens: within tumors, metformin shifts the balance away from immunosuppressive M2 macrophages and toward inflammatory, tumor-fighting M1 macrophages.¹¹ A review of metformin’s effects on the tumor immune microenvironment noted that it also reduces the activity of tumor-associated macrophages and cancer-associated fibroblasts, both of which normally help tumors evade the immune system.³⁵

How metformin “knows” to behave differently inside a tumor versus inside an inflamed joint or blood vessel remains an active research question. Part of the answer likely lies in the metabolic environment of the tissue: tumors are metabolically abnormal in ways that alter how immune cells respond to AMPK activation and energy stress. The practical implication for patients is that metformin’s anti-inflammatory profile does not appear to undermine anti-cancer immunity, and may even help it, though clinical evidence on cancer outcomes in metformin users is still mixed and should not be oversimplified.