
Metformin
Metformin is the usual first medicine for type 2 diabetes. It lowers HbA1c by about 1 percentage point on its own and rarely causes low blood sugar unless combined with insulin or a sulfonylurea. Long-term use can lower vitamin B12. Several supplements sold for blood sugar add a small extra lowering effect, and heavy alcohol use raises the risk of lactic acidosis.
Prescription medication: not a dietary supplement
This is an FDA-approved (or investigational) drug, not a supplement. It requires a prescription and medical supervision. The information below summarizes clinical-trial evidence for education only. It is not a recommendation to obtain or use it without a doctor.
Quick decision guide
May help most
Adults (and children 10 and older for immediate-release tablets) with type 2 diabetes whose kidney function allows it, prescribed and monitored by a clinician
Common dosing range
Set by your prescriber. Labeled starting doses are 500 mg twice daily or 850 mg once daily with meals (immediate release), or 500 mg once daily with the evening meal (extended release), raised slowly; label maximums are 2,550 mg a day (immediate release) and 2,000 mg a day (extended release)
When to expect effects
Blood sugar starts to fall within days to weeks; HbA1c reflects the change after about 3 months
Watch out for
Boxed warning for lactic acidosis (rare; higher risk with poor kidney function, heavy alcohol, liver disease, contrast dye scans, surgery and severe illness). Long-term use lowers vitamin B12 in some people: the label advises checking B12 every 2 to 3 years, so ask your doctor whether to check your level
What is it
Metformin is a prescription tablet that lowers blood sugar in type 2 diabetes. It belongs to a drug class called biguanides and has been used for decades; it is usually the first diabetes medicine people are offered, alongside diet and exercise. It comes as immediate-release tablets taken with meals, extended-release tablets taken once a day with the evening meal, and a liquid. Brand names have included Glucophage, Glucophage XR, Glumetza and Fortamet, but most people take a generic. Because many people on metformin also use "blood sugar support" supplements, and because metformin can lower vitamin B12 over time, supplement questions come up often.
Is it worth it for you?
Use this as a quick fit check, not a diagnosis.
Worth considering if…
Probably skip if…
Evidence at a glance
| Goal | Effect | Best fit | Time |
|---|---|---|---|
blood sugar (HbA1c) in type 2 diabetes Good Evidence | HbA1c about 1.1 points lower than placebo as monotherapy; 0.6 to 0.95 points as add-on | Adults with type 2 diabetes and adequate kidney function | About 3 months for the full HbA1c effect |
diabetes complications and death in overweight people with type 2 diabetes Good Evidence | 32 percent fewer diabetes-related endpoints and 36 percent lower all-cause death versus diet in UKPDS 34; pooled all-cause mortality relative risk 0.96 (95% CI 0.84 to 1.09) across trials | Overweight adults with newly diagnosed type 2 diabetes | Years (median follow-up 10.7 years) |
delaying type 2 diabetes in people with prediabetes Good Evidence | 31 percent fewer new diabetes cases at 2.8 years; 18 percent at 15 years (hazard ratio 0.82) | Overweight adults with prediabetes (raised fasting and post-meal glucose) | Seen within about 3 years |
eye, kidney and nerve complications in people with prediabetes Good Evidence | No difference: 13.0 vs 12.4 percent aggregate microvascular outcome | Not applicable | Not applicable |
blood sugar (HbA1c) in type 2 diabetes
- Effect
- HbA1c about 1.1 points lower than placebo as monotherapy; 0.6 to 0.95 points as add-on
- Best fit
- Adults with type 2 diabetes and adequate kidney function
- Time
- About 3 months for the full HbA1c effect
diabetes complications and death in overweight people with type 2 diabetes
- Effect
- 32 percent fewer diabetes-related endpoints and 36 percent lower all-cause death versus diet in UKPDS 34; pooled all-cause mortality relative risk 0.96 (95% CI 0.84 to 1.09) across trials
- Best fit
- Overweight adults with newly diagnosed type 2 diabetes
- Time
- Years (median follow-up 10.7 years)
delaying type 2 diabetes in people with prediabetes
- Effect
- 31 percent fewer new diabetes cases at 2.8 years; 18 percent at 15 years (hazard ratio 0.82)
- Best fit
- Overweight adults with prediabetes (raised fasting and post-meal glucose)
- Time
- Seen within about 3 years
eye, kidney and nerve complications in people with prediabetes
- Effect
- No difference: 13.0 vs 12.4 percent aggregate microvascular outcome
- Best fit
- Not applicable
- Time
- Not applicable
Evidence for 4 uses
AI-assisted evidence assessment. Talk to your doctor before relying on any single supplement.
blood sugar (HbA1c) in type 2 diabetes
Biomarker supportA 2012 meta-analysis of 35 trials lasting at least 12 weeks found that metformin on its own lowered HbA1c by 1.12 percentage points compared with placebo. Added to other diabetes pills it lowered HbA1c by 0.95 points, and added to insulin by 0.60 points. Higher doses lowered HbA1c more than lower doses without a significant increase in side effects. Results varied between trials (I-squared about 80 percent), but every analysis favored metformin.
Bottom line: Reliably lowers HbA1c by about 1 point on its own.
diabetes complications and death in overweight people with type 2 diabetes
Disease adjunctIn UKPDS 34, 753 overweight adults (mean age 53) with newly diagnosed type 2 diabetes were randomized to metformin-based intensive control or mainly diet alone and followed for a median of 10.7 years. Metformin lowered any diabetes-related endpoint by 32 percent, diabetes-related death by 42 percent and death from any cause by 36 percent, with less weight gain and fewer low blood sugar episodes than insulin or sulfonylureas. However, in a supplementary part of the same study, adding metformin early to a sulfonylurea was linked to more diabetes-related deaths. A 2017 meta-analysis of 13 trials, in which UKPDS supplied most of the data, found effects that favored metformin but none that were statistically significant, and the label states there is no conclusive evidence that metformin reduces heart and blood vessel disease.
Bottom line: One landmark trial showed fewer complications and deaths; pooled trials are less certain.
Evidence is mixed
A 2017 meta-analysis of 13 trials found no statistically significant reduction in death, heart attack or stroke, and the supplementary UKPDS trial found more diabetes-related deaths when metformin was added early to a sulfonylurea.
delaying type 2 diabetes in people with prediabetes
Disease adjunctThe Diabetes Prevention Program randomized 3,234 US adults (mean age 51, mean BMI 34) with raised fasting and post-meal glucose to placebo, metformin 850 mg twice daily, or an intensive lifestyle program. Over 2.8 years, diabetes developed in 11.0, 7.8 and 4.8 people per 100 per year: metformin cut new cases by 31 percent and lifestyle change by 58 percent, so lifestyle worked better. Over 15 years of follow-up the reduction with metformin was 18 percent, and the gap narrowed over time. Metformin is not FDA-approved for preventing diabetes; this use is a decision for your doctor.
Bottom line: Delays diabetes in prediabetes, but less than lifestyle change.
eye, kidney and nerve complications in people with prediabetes
No benefit foundIn the 15-year Diabetes Prevention Program Outcomes Study, the combined rate of kidney, eye and nerve damage at the end of the study was 13.0 percent with metformin and 12.4 percent with placebo, not significantly different, even though metformin had reduced new diabetes cases. Across the study, people who did not develop diabetes had 28 percent fewer complications than people who did.
Bottom line: Fewer diabetes diagnoses did not translate into fewer small-vessel complications over 15 years.
How it works
How to take it
What to track
Bottom line: Take it with meals, keep alcohol moderate, and ask about B12 and kidney checks.
Safety
Know the common side effects, key cautions, and who should avoid it.
Common side effects
Serious risks
Boxed warning: lactic acidosis. Rare but can be fatal. Early signs are vague: feeling unwell, muscle pain, trouble breathing, unusual sleepiness and stomach pain. Risk is higher with kidney impairment, age 65 or older, contrast dye scans, surgery, low-oxygen states such as acute heart failure, heavy alcohol use, liver disease and some medicines. Get urgent care if these symptoms appear.
Vitamin B12 deficiency. In 29-week label trials about 7 percent of people developed below-normal B12. In a 4.3-year randomized trial in 390 adults in the Netherlands with type 2 diabetes who used insulin, metformin 850 mg three times daily lowered B12 by 19 percent versus placebo and deficiency was 7.2 percentage points more common. Low B12 can cause anemia and nerve damage.
Low blood sugar when combined with insulin or a sulfonylurea; the label says those doses may need to be lowered.
Liver injury (cholestatic, hepatocellular or mixed) has been reported after approval.
Who should avoid it
- People with severe kidney impairment (eGFR below 30); starting it is not recommended with eGFR 30 to 45
- People with metabolic acidosis, including diabetic ketoacidosis, or a known allergy to metformin
- People with liver disease, or mitochondrial diseases such as MELAS or MIDD, where the label advises against use
Pregnancy & breastfeeding
The label says published studies have not shown a clear link between metformin and birth defects or miscarriage, but the data cannot rule out a risk. Poorly controlled diabetes in pregnancy is itself risky for mother and baby. The label also notes that metformin may cause ovulation in some premenopausal women who were not ovulating, which may lead to an unintended pregnancy. Discuss pregnancy plans with your prescriber.
Interactions
Metformin does not block a B12 supplement from working, but long-term use lowers B12 absorption from food. In a 4.3-year trial in 390 adults with type 2 diabetes using insulin, deficiency was 7.2 percentage points more common than with placebo, and in the Diabetes Prevention Program each year of metformin raised the odds of deficiency. The label advises measuring B12 every 2 to 3 years; ask your doctor whether to check your level.
The label says alcohol increases metformin's effect on lactate and lists excessive alcohol intake as a risk factor for lactic acidosis, and it warns patients against excessive drinking. Ask your doctor or pharmacist what amount is safe for you, especially if you have kidney or liver problems.
MK-677 raises growth hormone, which makes the body less sensitive to insulin. In a 2-year trial in healthy adults aged 60 to 81, 25 mg a day raised fasting blood sugar by an average of 5 mg/dL and lowered insulin sensitivity, working against what metformin does. It is not an approved medicine; tell your prescriber if you take it so your blood sugar can be watched.
Glucomannan forms a thick gel that slowed absorption of the diabetes drug glibenclamide in volunteers, but those volunteers took glibenclamide, not metformin, and in lab tests at stomach and gut acidity it did not bind metformin. Taking it at a different time from metformin, with plenty of water, and watching your glucose when you start it are reasonable precautions to review with your doctor or pharmacist.
Viscous fibers can slow metformin absorption: guar gum, a similar gel-forming fiber, slowed metformin absorption over 6 hours in 6 volunteers. In a small 12-week psyllium study, the 5 people on metformin alone saw HbA1c rise slightly while those on no diabetes medicine saw it fall. Evidence is very limited; tell your prescriber or pharmacist you use psyllium and ask about spacing it from your dose.
These medicines can make the blood more acidic, and the label says combining them with metformin may increase the risk of lactic acidosis. It advises more frequent monitoring; make sure your prescribers know you take both.
These drugs block the kidney transporters that remove metformin, which can raise metformin levels and the risk of lactic acidosis. The label asks prescribers to weigh the benefits and risks of using them together.
Metformin rarely causes low blood sugar alone, but combined with insulin or a sulfonylurea it can. The label says a lower dose of insulin or the sulfonylurea may be needed; your prescriber decides.
Fiber in a meal may slow metformin absorption a little, based on studies of gel-forming fibers. Metformin is meant to be taken with meals, fiber-rich ones included, and no source sets a gap for ordinary food. Concentrated fiber supplements are a separate question to raise with your prescriber.
Alpha-lipoic acid lowers blood sugar slightly by itself. A 2022 meta-analysis of 16 trials (1,035 people with type 2 diabetes) found HbA1c about 0.3 points lower, below the level judged clinically important. The added effect on metformin is small; tell your diabetes team, especially if you also take insulin or a sulfonylurea.
Trials of chromium in type 2 diabetes disagree: a 2020 meta-analysis found lower fasting glucose and HbA1c with very large differences between studies, while an earlier meta-analysis called the data inconclusive and found no effect in people without diabetes. Any added effect on metformin is uncertain; tell your prescriber and watch your readings.
A meta-analysis of 24 trials in type 2 diabetes found cinnamon supplements lowered fasting glucose and HbA1c, so they may add a little to metformin's effect. Kitchen amounts are not a concern; if you take it as a supplement, tell your prescriber and monitor your glucose.
This is an additive (synergy) effect rather than a harmful one: in a 2018 meta-analysis of 10 small trials (490 people), ginger capsules or tablets of 1 to 3 g a day lowered HbA1c and fasting glucose, so ginger may add modestly to metformin. The trials did not test ginger tea. If you take concentrated ginger and also use insulin or a sulfonylurea, watch for lows and review it with your doctor or pharmacist.
Documented interactions
Evidence-graded pair pages with sources, dosing notes, and timing guidance, a complement to the narrative section above.
Warnings (9)
+ alcohol
highAlcohol raises the risk of metformin-associated lactic acidosis and can trigger delayed low blood sugar, especially with heavy drinking or when drinking without food.
+ vitamin b12
highLong-term metformin use can reduce vitamin B12 absorption, sometimes enough to cause deficiency.
+ glucomannan
moderateGlucomannan is a highly viscous soluble fiber that swells in the gut. It slowed absorption of another diabetes drug, glibenclamide, in 9 volunteers, and guar gum, a similar gel-forming fiber, slowed metformin absorption in 6 volunteers. In lab tests at stomach and gut acidity, glucomannan did not bind metformin, so any effect on metformin in people is unproven. Glucomannan also has its own modest glucose-lowering effect that may add to metformin's, so spacing the two apart and watching your blood sugar is sensible.
+ mk-677
moderateMK-677 raises growth hormone, and growth hormone reduces the body's sensitivity to insulin. Across controlled trials this shows up as rising fasting blood sugar and rising HbA1c. In a two-year randomised trial, sixteen of forty-three participants taking MK-677 moved into the impaired fasting glucose range, against two of twenty-two on placebo, and insulin sensitivity fell measurably. A 2025 open-label pilot study, published as a conference abstract, saw HbA1c rise compared with a historical placebo group, and two participants dropped out because of high blood sugar. This works directly against what metformin is prescribed to do.
Choosing a product
What to look for on the label, and what to be skeptical of.
Look for…
Be skeptical of…
Frequently asked questions
Should I take vitamin B12 with metformin?⌄
Long-term metformin lowers B12 in some people: in a 4.3-year trial, deficiency was about 7 percentage points more common than with placebo. The label advises checking B12 every 2 to 3 years. Ask your doctor whether to check your level; they can tell you whether you need a supplement and how much.
Can I drink alcohol on metformin?⌄
The label warns against excessive alcohol intake because alcohol increases metformin's effect on lactate and raises the risk of lactic acidosis, a rare but serious side effect. Binge drinking and drinking with kidney or liver problems are the biggest concerns. Ask your doctor or pharmacist what is safe for you.
Can I take cinnamon or chromium with metformin?⌄
Cinnamon and chromium may add a small extra glucose-lowering effect; trials are mixed, especially for chromium. On metformin alone, lows are unlikely, but the risk rises if you also take insulin or a sulfonylurea. Tell your prescriber about any blood sugar supplement and watch your readings when you start one.
Does fiber or psyllium interfere with metformin?⌄
Possibly a little. A gel-forming fiber (guar gum) slowed metformin absorption in a small study, and a tiny psyllium study saw HbA1c rise slightly in people on metformin alone. Fiber-rich meals are fine, since metformin is taken with food. For concentrated fiber supplements, ask your prescriber or pharmacist about spacing them from your dose.
Why does metformin upset my stomach?⌄
Diarrhea, nausea and gas are the most common side effects because metformin acts in the gut. In label trials diarrhea was much less common with the extended-release form (10 percent) than with immediate release (53 percent). Taking it with meals helps. If side effects persist, talk to your prescriber rather than stopping it.
References by claim
blood sugar (HbA1c) in type 2 diabetes
Hirst et al., 2012 · PMC (2012) link
diabetes complications and death in overweight people with type 2 diabetes
delaying type 2 diabetes in people with prediabetes
Safety
Interaction: MK-677 (ibutamoren)
Nass et al., 2008 · PMC (2008) link
Interaction: Glucomannan (konjac fiber)
Interaction: Alpha-lipoic acid
Jibril et al., 2022 · PMC (2022) link
Interaction: Chromium
Interaction: Cinnamon
Moridpour et al., 2024 · PubMed (2024) link
Interaction: Ginger tea
Zhu et al., 2018 · PMC (2018) link
Track Metformin with Pilora
Set up dose reminders, check interactions, and join the community in the Pilora app.
Disclaimer: This page summarizes published clinical-trial data for educational purposes and is not medical advice or a recommendation to use this prescription medication. Dosing, eligibility, and monitoring must be decided by a licensed prescriber.
