When the UK health authority warned doctors in 2022 that a common diabetes medicine called metformin might lower vitamin B12 levels, researchers wanted to know if doctors actually started testing patients more often. They looked at medical records from nearly 150,000 people and found that yes, doctors did test more patients for B12 after the warning—but only a little bit more. Interestingly, doctors still weren’t documenting B12 deficiency cases any better than before, suggesting the warning helped somewhat but didn’t completely change how doctors handle this issue.

The Quick Take

  • What they studied: Whether a 2022 official safety warning about metformin and vitamin B12 actually changed how often doctors tested diabetic patients for B12 deficiency and how they recorded these findings.
  • Who participated: Nearly 148,000 people’s medical records from three healthcare networks in Southwest England, including both people taking metformin for diabetes and people in the general population, tracked from 2017 to 2024.
  • Key finding: After the 2022 safety warning, B12 testing increased from 34.5% to 38.2% in metformin patients—a modest but real increase. However, the number of doctors actually documenting B12 deficiency stayed the same, suggesting the warning only partially worked.
  • What it means for you: If you take metformin for diabetes, there’s a slightly better chance your doctor will test your B12 levels now than before 2022. However, this study suggests there’s still room for improvement in how consistently doctors monitor and record B12 problems in diabetic patients.

The Research Details

Researchers looked back at medical records from 2017 to 2024 to see what changed after June 2022, when the UK’s medicine safety agency (MHRA) officially warned doctors that metformin could cause low B12 levels. They compared two groups: people taking metformin and people in the general population. They counted how many people got B12 tests and how many were diagnosed with B12 deficiency before and after the warning.

The study used electronic medical records from three healthcare networks in Southwest England, which gave them access to nearly 148,000 patient records. They used a statistical test called Welch’s t-test to figure out if the changes they saw were real or just random chance. They considered a result “real” if there was less than a 5% chance it happened by accident.

This approach is useful because it shows what actually happens in real doctor’s offices, not in a controlled research setting. By comparing metformin users to the general population, researchers could see if the warning specifically changed how doctors treated diabetic patients differently.

Understanding whether safety warnings actually change what doctors do is important. A warning is only helpful if doctors read it and act on it. This study helps show whether the MHRA’s effort to get doctors to test more patients actually worked in real-world practice. It also reveals a gap: even though doctors tested more, they didn’t document their findings better, which could mean patients aren’t getting proper follow-up care.

This study is strong because it used a large number of real patient records from actual doctor’s offices, not a small group of volunteers. The researchers used proper statistical methods to check if their findings were real. However, the study only looked at two years of data after the warning (2022-2024), which isn’t very long to see if changes stick around. The study also couldn’t explain why doctors didn’t document B12 deficiency better—it just showed that they didn’t.

What the Results Show

After the 2022 safety warning, B12 testing in metformin patients increased from 34.5% to 38.2%—an increase of about 3.7 percentage points. This increase was statistically significant, meaning it was unlikely to have happened by chance. Interestingly, B12 testing also increased in the general population (from 12.2% to 14.7%), though this increase was smaller. This suggests the warning may have had some effect, but it also shows that testing increased everywhere, not just for diabetic patients.

The more surprising finding was about documentation. Among metformin patients, the rate of doctors actually coding (recording) a B12 deficiency diagnosis stayed flat at 0.25% before and after the warning. In the general population, this rate actually decreased slightly from 0.072% to 0.060%. This means that even though doctors were testing more, they weren’t finding or recording B12 deficiency any more often.

These results suggest a disconnect: the warning got doctors to order more tests, but it didn’t lead to better identification or documentation of actual B12 problems. This could mean several things: doctors might not be following up on test results properly, they might not be recording findings in a way the system can track, or they might be finding low B12 but not considering it serious enough to document.

The study found that the increase in testing was modest in both groups, suggesting that while the warning had an effect, it wasn’t dramatic. The fact that testing increased in the general population too (even though they weren’t specifically mentioned in the warning) suggests that other factors beyond the MHRA warning might have influenced testing patterns. This could include increased awareness of B12 deficiency in general or other healthcare trends happening at the same time.

This study builds on decades of research showing that metformin can lower B12 levels. Previous studies have shown that 10-30% of long-term metformin users develop low B12 levels, but doctors haven’t consistently tested for this. The 2022 MHRA warning was meant to address this gap. This new study shows that the warning did have some effect—more testing happened—but it confirms what researchers suspected: simply issuing a warning doesn’t automatically change how doctors practice medicine.

The biggest limitation is that the study only looked at two years of data after the warning (2022-2024), which is a short time to see if changes are permanent. The study also couldn’t explain why doctors didn’t document B12 deficiency better—it just showed that they didn’t. The researchers couldn’t see why testing increased in the general population or whether the tests that were done actually found B12 problems. Additionally, the study only looked at one region of England, so results might be different in other parts of the UK or other countries.

The Bottom Line

If you take metformin for type 2 diabetes, ask your doctor about B12 testing, especially if you’ve been on metformin for several years. There’s now a slightly better chance your doctor will offer this test, but you shouldn’t assume it will happen automatically. If you experience symptoms like fatigue, numbness, or tingling, mention these to your doctor and ask about B12 testing. (Confidence level: Moderate—based on this study plus decades of previous research showing metformin can lower B12.)

This matters most for people who have taken metformin for more than a few years, as B12 problems develop over time. It’s less relevant for people just starting metformin or those taking it for a short time. Doctors and healthcare administrators should care about this study because it shows that safety warnings alone might not be enough to change practice—they may need additional support or training to implement new guidelines.

B12 deficiency develops slowly in metformin users, usually over several years. If you start getting tested now, it could take months to years to see if you develop low B12. If you do have low B12, it typically takes weeks to months of treatment (usually B12 supplements or injections) to feel better, though some symptoms like nerve damage might take longer to improve.

Want to Apply This Research?

  • If you use a diabetes or health app, track when you get B12 tests and the results. Set a reminder to ask your doctor about B12 testing every 1-2 years if you’re on metformin, and log the date and result each time.
  • Use your app to set an annual reminder to discuss B12 testing with your doctor at your next appointment. If your doctor hasn’t mentioned it, bring it up yourself. You could also track any symptoms like unusual tiredness or tingling that might suggest low B12.
  • Create a long-term tracking log in your app that shows your B12 test dates and results over time. This helps you and your doctor spot trends. If results start dropping, your doctor can intervene earlier. You could also track energy levels and symptoms alongside test results to see if they correlate.

This study describes what happened in real doctor’s offices after a safety warning, but it doesn’t prove that the warning caused all the changes observed. If you take metformin, do not start or stop any medications or supplements without talking to your doctor first. B12 deficiency can have serious health effects if left untreated, but it’s also treatable. This article is for educational purposes and should not replace advice from your healthcare provider. If you have symptoms of B12 deficiency (such as fatigue, numbness, tingling, or memory problems), contact your doctor for proper evaluation and testing.