What is metformin and how has it been used?
Metformin is an inexpensive oral drug that has been a cornerstone of type 2 diabetes care for decades. Doctors routinely prescribe it as a first-line medication for people with type 2 diabetes. It lowers blood sugar, helps the body handle glucose more efficiently, and has been linked to reduced cardiovascular risk and lower mortality in large studies such as the UK Prospective Diabetes Study.
Even though metformin is widely used, its exact mechanism is still not fully understood. Clamp studies in type 2 diabetes show that it changes glucose metabolism in both muscle and liver. It also raises levels of growth differentiation factor 15, or GDF15, a stress-regulated hormone that is involved in appetite suppression, body weight, and glucose control.
Because it is cheap, generally safe, and appears to protect the heart in type 2 diabetes, metformin has often been used off label in people with type 1 diabetes. Many clinicians hoped that adding metformin to insulin would reduce insulin resistance, lower insulin requirements, and improve long term cardiovascular outcomes.
Why insulin resistance matters in type 1 diabetes
Type 1 diabetes is an autoimmune disease. The immune system destroys the insulin-producing beta cells in the pancreas, forcing people to depend on lifelong insulin replacement. Even with modern insulin therapy, cardiovascular disease remains the leading cause of death in type 1 diabetes, and traditional risk factors do not fully explain that risk.
Insulin resistance is a key metabolic problem in this group. In type 1 diabetes, insulin is delivered under the skin instead of being released directly into the portal vein that leads to the liver. This disturbed portal-peripheral insulin gradient appears to drive resistance in the liver, muscle, and adipose tissue. Studies have linked insulin resistance to cardiovascular risk, but it has not been clear whether muscle or liver resistance matters most.
The new research shows that adults with type 1 diabetes have multi-tissue insulin resistance even when their body mass index and visceral fat are similar to people without diabetes. At baseline in this study:
- Liver insulin resistance was higher, with endogenous glucose production 64 percent higher than in control participants.
- Muscle insulin resistance was higher, with glucose infusion rate during the high-dose clamp 29 percent lower than in controls.
- Adipose tissue was more resistant, with higher non-esterified fatty acids during the low-dose phase, showing impaired suppression of lipolysis.
Muscle insulin resistance in particular was strongly linked with cardiovascular risk markers such as arterial stiffness, inflammatory markers, liver enzymes, and low-density lipoprotein cholesterol. In people with type 1 diabetes, it also related to HbA1c, insulin dose, and overnight glucose levels, suggesting it may be a key driver of vascular complications.
Who did this new research and how was the study run?
The new trial, published in Nature Communications, was led by Jennifer R. Snaith, Jerry R. Greenfield, and colleagues in Australia. The work was conducted at a single site as part of the Insulin Resistance in Type 1 Diabetes Managed with Metformin study, known as INTIMET.
The researchers first carried out a cross-sectional study of 40 adults with type 1 diabetes and 20 adults without diabetes. All were about 37 years old on average, about 60 percent male, and had similar body mass index values. The type 1 diabetes group had lived with the disease for around 23 years, had an HbA1c of 7.5 percent, and used either multiple daily injections or insulin pumps.
Insulin resistance in liver, muscle, and adipose tissue was measured using a three-stage hyperinsulinemic-euglycemic clamp with isotope-labeled glucose, the gold standard technique.
The team then ran a phase 3 randomized, double-blind, placebo-controlled trial. Forty adults with type 1 diabetes were randomly assigned to metformin or placebo. After titration, participants took up to 1,500 milligrams of metformin a day for 26 weeks, or matching placebo tablets.
The primary goal was to see whether adding metformin to insulin would reduce hepatic insulin resistance, measured as the change in endogenous glucose production during the low-dose step of the clamp. Secondary goals included changes in muscle and adipose insulin sensitivity, insulin dose, glucose control, visceral fat mass, cardiovascular markers, and GDF15 levels.
Thirty-seven of the 40 participants completed the 26-week intervention, and adherence to medication was high in both groups.
What did the trial actually find?
After six months of treatment, metformin did not improve insulin resistance in adults with type 1 diabetes. The main tissue-level findings were:
- Liver insulin resistance: There was no significant difference in the change in endogenous glucose production between the metformin and placebo groups after 26 weeks. The mean difference was 0.2 micromoles per kilogram of fat-free mass per minute, with a 95 percent confidence interval from minus 0.4 to 0.8 and a p value of 0.53.
- Muscle insulin resistance: Glucose infusion rate increased over time in both groups, but the difference between metformin and placebo was not significant. The treatment by time interaction for glucose infusion rate was not different between groups.
- Adipose insulin resistance: There was no significant difference between treatment groups in the change in non-esterified fatty acid levels over the 26 weeks.
In short, metformin did not improve insulin sensitivity in liver, muscle, or fat tissue when measured directly by clamp.
However, there were some important secondary results:
- Metformin significantly lowered total daily insulin dose compared with placebo, with an estimated treatment difference of minus 0.10 units per kilogram per day and a p value of 0.0008. This effect remained significant after adjustment for multiple comparisons and did not depend on age, sex, baseline HbA1c, body mass index, or insulin delivery method.
- HbA1c and continuous glucose monitoring metrics such as time in range, time below range, time above range, and glucose variability did not differ between groups over 26 weeks.
- There was a trend toward lower visceral adipose tissue mass in the metformin group, but this did not remain significant after correcting for multiple comparisons.
- Metformin markedly increased circulating GDF15 levels compared with placebo, consistent with its known biological effects in other populations.
- Metformin was well tolerated. There were no episodes of diabetic ketoacidosis or severe hypoglycemia in either group. Gastrointestinal side effects were actually more frequent in the placebo group.
Why did metformin fail to improve tissue-level insulin resistance?
This is the first randomized controlled trial to test the effect of metformin on directly measured insulin resistance in adults with type 1 diabetes. Earlier studies in adolescents had suggested that metformin might improve peripheral insulin sensitivity, but those trials did not see changes in hepatic glucose production. The authors suggest that puberty and growth-related factors may explain why adolescents responded differently than adults.
The discussion in this study points to several factors that may blunt metformin’s impact on tissue insulin resistance in adults with type 1 diabetes:
- Muscle insulin resistance appears to be strongly influenced by chronic hyperglycemia and by the way insulin is delivered. Subcutaneous injections or pump infusions send insulin straight into the peripheral circulation, bypassing the liver and altering the normal balance between portal and peripheral insulin levels.
- Long term peripheral hyperinsulinemia may disrupt insulin signaling and contribute to defects in glucose disposal, endothelial function, and body composition.
- Liver fat was not higher in the type 1 diabetes group and did not relate to liver insulin resistance, which differs from many type 2 diabetes patterns. This suggests that hepatic insulin resistance may be secondary to peripheral defects rather than a direct fat-driven process.
The researchers also note that metformin’s ability to reduce insulin dose without improving clamp-measured insulin resistance may be driven by other mechanisms, such as effects on the gut. Prior work in other settings has shown that metformin increases intestinal glucose uptake, alters transporter expression in enterocytes, affects bile acid absorption, incretin hormones, GDF15, and the gut microbiome. The authors speculate that GDF15 may help create an “intestinal glucose sink,” allowing lower insulin doses without classic improvements in peripheral insulin sensitivity.
What does this mean for using metformin in diabetes treatment?
The authors are clear in their main conclusion. After 26 weeks of therapy, metformin did not improve liver, muscle, or adipose tissue insulin resistance in adults with type 1 diabetes. They state that their results do not support prescribing metformin to reduce hepatic insulin resistance in this population.
At the same time, the trial complicates the picture rather than closing the door on metformin entirely. Several findings keep metformin in the conversation:
- Metformin consistently reduced total daily insulin dose by about 0.1 units per kilogram per day relative to placebo.
- Prior long term follow up data from the Diabetes Control and Complications Trial and its Epidemiology of Diabetes Interventions and Complications extension found that each 0.1 units per kilogram per day increase in insulin dose predicted a 6 percent rise in cardiovascular disease risk over 30 years, although much of this association was explained by traditional risk factors. This raises the possibility that insulin sparing could have benefits over time.
- The REMOVAL trial, which studied metformin in older adults with type 1 diabetes and high cardiovascular risk, did not find a significant reduction in its primary endpoint of mean carotid intima-media thickness, but a subgroup of never-smokers did show benefit. That result, along with the current trial’s insulin dose reduction and GDF15 increase, suggests that metformin may still have pleiotropic cardiovascular effects in type 1 diabetes, even if it does not directly improve insulin resistance.
For now, the clinical relevance of lowering insulin dose without changing tissue-level insulin resistance remains uncertain. Insulin monotherapy continues to be the mainstay of treatment in type 1 diabetes, and the study authors emphasize that multi-tissue insulin resistance is a prominent defect that still needs better targeted therapies.
What are researchers and clinicians saying?
The investigators behind the INTIMET study argue that their work fills a major gap. No previous trial had directly measured changes in insulin resistance with metformin in adults with type 1 diabetes using gold-standard clamp techniques. Their data challenge the common assumption that a lower insulin dose automatically reflects improved insulin sensitivity.
They also highlight important limitations. The sample size was modest, the cohort was largely Caucasian, and people with HbA1c above 9.5 percent were excluded. These factors may limit how widely the results apply and may have reduced the power to detect smaller changes in muscle insulin resistance or other secondary outcomes.
Even with these caveats, the main message is firm. In adults with type 1 diabetes across a broad range of insulin sensitivities, six months of metformin did not improve tissue-specific insulin resistance. As the authors conclude, their findings do not support prescribing metformin with the goal of reducing insulin resistance in type 1 diabetes, but they do point toward other potential roles for metformin, especially insulin sparing and possible cardiovascular protection, that will require further study.
