Can metformin reduce cancer risk in diabetic patients?

What does the current evidence say about metformin and cancer risk?

The answer is not a simple yes or no. The strongest evidence points to a protective effect for certain cancers, particularly colorectal, liver, and cervical cancers, but the picture is complicated by conflicting results for other cancers and by the quality of the studies. A large, well-designed study from the UK that emulated a clinical trial found that metformin did not meaningfully reduce overall cancer risk [4]. This study is important because it corrected for common biases in observational research that can make metformin look more protective than it really is. However, many other large, population-based studies—which are more numerous—do find a benefit.

For colorectal cancer, the evidence is fairly consistent. A 2024 study from the Southern Community Cohort Study, which included a high proportion of Black and low-income participants, found that metformin users had a 29% lower risk of colorectal cancer (hazard ratio [HR] 0.71) [2]. This benefit was especially strong for rectal cancer (51% lower risk) and in obese individuals (41% lower risk) [2]. For liver cancer, a meta-analysis of five studies estimated a 62% reduction in risk [11]. For cervical cancer, a large Korean study found a 22% lower risk, with the benefit increasing with higher doses and longer use—up to an 86% lower risk for those on the highest doses [1].

The evidence for breast and lung cancers is more mixed. A meta-analysis of 15 studies found no statistically significant reduction in breast cancer risk among metformin users [8]. However, a large prospective study found that long-term metformin use (10+ years) was linked to a 38% lower risk of estrogen receptor-positive breast cancer, but a 74% increased risk of triple-negative breast cancer [7]. For lung cancer, one study found no overall benefit, but did find a 43% lower risk among never-smokers who used metformin [10]. These mixed results suggest that the effect of metformin may depend on the specific biology of the cancer and the patient's characteristics.

Does the dose or duration of metformin use matter?

Yes, several studies show that the protective effect of metformin is stronger with higher cumulative doses and longer duration of use. For cervical cancer, the risk reduction jumped from 22% overall to 86% for those who took a high cumulative dose (over 1,200,000 mg) and 86% for those who used it for an extended period (over 2,000 days) [1]. Similarly, a Korean nationwide study found that patients who started metformin early and took it consistently had a 37% lower risk of cancer overall, compared to those who did not use it [6]. This suggests that consistent, long-term use is key.

Other factors also play a role. The benefit of metformin for colorectal cancer was seen primarily in obese individuals (BMI ≥30), with a 41% risk reduction, but no benefit was seen in non-obese participants [2]. For lung cancer, the benefit was only seen in never-smokers, not in current or former smokers [10]. For breast cancer, the protective effect was enhanced in women who also used statins (cholesterol-lowering drugs) [9]. These findings suggest that metformin's cancer-preventive effects may be most pronounced in specific subgroups, and that it may work through mechanisms related to obesity, inflammation, and insulin resistance.

What are the important caveats and limitations of this research?

The most important caveat is that most of the evidence comes from observational studies, not randomized controlled trials (RCTs). Observational studies can be biased by 'confounding by indication'—meaning that people who take metformin may be healthier or have better access to care than those who don't, which could make the drug look more effective than it is. The UK study that emulated a trial found that when these biases were corrected, the apparent protective effect of metformin disappeared [4]. This is a major reason for caution.

Another limitation is that many studies do not adequately control for other diabetes medications, such as insulin or sulfonylureas, which may themselves increase cancer risk. Also, the studies vary widely in population (Korean, US, UK, Taiwanese), cancer type, and how they define metformin use (e.g., 'ever use' vs. 'current use' vs. 'duration of use'), making direct comparisons difficult. For example, the study on bone cancer found a benefit only in men over 60 [5], while the study on gestational diabetes found no effect on cancer outcomes because it was focused on pregnancy [3]. Finally, some of the positive findings, like the 62% reduction in liver cancer, come from a meta-analysis of studies published in 2012, which may be dated [11]. More recent, high-quality RCTs are needed to confirm these findings.

About These Sources

This answer is built on 11 peer-reviewed studies — published from 2012 to 2024, 2 from 2024 or later, 5 in Q1 journals, collectively cited 264 times — selected as the most relevant from 15 studies that passed quality screening, drawn from 71 papers retrieved from a database of over 500 million.

Sources used in this answer

Related research questions