What Creatinine Level Means You Should Stop Metformin?

There is no single creatinine number that reliably tells you when to stop metformin, because guidelines no longer use serum creatinine alone to make that call. Until 2016, the U.S. FDA package insert for metformin listed fixed creatinine cutoffs: 1.5 mg/dL or higher for men and 1.4 mg/dL or higher for women. Those thresholds have been replaced with a kidney-function measure called eGFR (estimated glomerular filtration rate), which accounts for age, sex, and body size in addition to creatinine. The consensus across major guidelines is that metformin should be stopped when eGFR drops below 30 mL/min/1.73 m², and the dose should be reduced in the range between 30 and 45.

Why the Old Creatinine Cutoffs Were Abandoned

Serum creatinine is a waste product generated by muscle metabolism. Your kidneys filter it out, so when your kidneys struggle, creatinine rises. The problem is that creatinine levels depend heavily on factors that have nothing to do with kidney health. A muscular young man can walk around with a creatinine of 1.3 mg/dL and have perfectly fine kidneys, while a thin elderly woman with a creatinine of 1.0 mg/dL can have seriously impaired kidney function. One study of elderly patients found that a creatinine threshold of 1.7 mg/dL detected only about 13% of patients who actually had kidney failure, meaning nearly 9 out of 10 cases were missed. The test performed inconsistently across sexes: women with severe kidney failure had significantly lower creatinine levels than men with the same degree of impairment.

The core issue is that a creatinine number only makes sense once you factor in the person it came from. A creatinine of 1.4 mg/dL in a 25-year-old man corresponds to a very different level of kidney function than the same number in a 75-year-old woman. The eGFR calculation does this adjustment automatically. That is why every major diabetes and nephrology society now frames metformin decisions around eGFR rather than raw creatinine.

The eGFR Thresholds That Guide Metformin Use

Current guidelines break kidney function into tiers for metformin prescribing. While minor variations exist between organizations, the broad consensus follows a three-tier structure:

eGFR 45 or above: Metformin can be used safely at standard doses. This covers the vast majority of people with type 2 diabetes.
eGFR 30 to 44: Metformin should generally not be newly started. If someone is already taking it and doing well, the dose should be capped at 1,000 mg per day, and kidney function should be monitored more frequently.
eGFR below 30: Metformin is contraindicated and should be stopped.

These thresholds are endorsed by the Korean Diabetes Association and Korean Society of Nephrology consensus, and they align closely with guidance from the FDA, the American Diabetes Association, and other international bodies.² A practical review of metformin optimization reinforces this: when eGFR is between 30 and 60, the dose should be reconsidered and patients should receive education about when to temporarily hold the drug during illness. Metformin should be discontinued when eGFR falls below 30.³

If you only have a creatinine result and want a rough sense of where you stand, you can ask your doctor to calculate your eGFR, or use one of the freely available online calculators (the CKD-EPI equation is the current standard). But this is genuinely something your care team should be tracking, because the math involves variables you may not have at hand.

Lactic Acidosis and Why the Threshold Exists

The reason metformin is stopped at low kidney function is the risk of lactic acidosis, a condition where lactic acid builds up in the blood faster than the body can clear it. Metformin is cleared almost entirely by the kidneys. When kidney function drops, the drug accumulates, and at high enough concentrations it interferes with how cells produce energy. Specifically, metformin inhibits a step in the mitochondrial energy chain, which pushes cells toward producing lactate instead of fully burning glucose for energy.⁴ When the liver’s ability to process that lactate is also overwhelmed, dangerous acidosis can result.⁵

That said, the absolute risk of metformin-associated lactic acidosis (MALA) is genuinely rare. A large systematic review pooling data from 194 studies found zero cases of fatal or nonfatal lactic acidosis across roughly 37,000 patient-years of metformin use. The estimated upper limit for the true rate was about 8 cases per 100,000 patient-years, which was no different from the rate in people taking other diabetes drugs.⁶ That finding has led some researchers to argue that the fear of MALA has been disproportionate to its actual frequency, and that overly cautious creatinine cutoffs historically denied metformin’s benefits to many people who could have used it safely.

However, the picture changes when you look at critically ill patients. In a retrospective study of ICU patients with type 2 diabetes, the rate of lactic acidosis was higher in metformin users than in non-users (about 6% versus 4% after statistical matching), and metformin exposure was independently associated with nearly double the odds of lactic acidosis. Even so, in-hospital mortality was not significantly different between the two groups.⁷ A separate analysis estimated the real-world incidence at roughly 47 cases per 100,000 patient-years, several times higher than older estimates, likely because many patients were using metformin in the presence of other risk factors for lactic acidosis. Most of the MALA cases in that study had preexisting risk factors, and about 80% had kidney failure at the time of admission.⁸

The takeaway is that metformin-associated lactic acidosis is rare but not mythical. It tends to happen when kidney function is already compromised or when another acute illness (severe dehydration, sepsis, heart failure) piles on top. The eGFR threshold of 30 exists not because MALA is common at that point, but because the margin of safety shrinks considerably once kidneys can no longer clear the drug effectively.

When to Temporarily Hold Metformin

Even if your kidney function is fine on a routine lab draw, there are situations where metformin should be paused temporarily. The most common is the “sick day” scenario: any illness involving vomiting, diarrhea, high fever, or poor fluid intake can acutely drop your kidney function and impair the body’s ability to clear metformin. A randomized trial of a sick-day protocol in patients with chronic kidney disease instructed patients to stop metformin (along with certain other drugs) for up to 48 hours when they became acutely ill, resuming only when the illness resolved.⁹

The other widely recognized scenario is receiving iodinated contrast dye for a CT scan or certain cardiac procedures. Contrast agents can temporarily reduce kidney function, and if metformin is still circulating at that point, it may accumulate. Standard practice at most hospitals is to hold metformin on the day of the procedure and restart it 48 hours later once kidney function has been confirmed stable. If your eGFR is already in the cautious zone (below 45 or so), your doctor may hold it even longer or recheck labs before restarting.

The broader principle here is that metformin safety is not just about your baseline kidney function. It is about what happens to your kidneys under stress. A person with an eGFR of 50 at their last checkup can temporarily drop to 25 during a bad stomach flu. That transient dip is where the danger lies.

Benefits of Staying on Metformin With Moderate Kidney Disease

Part of the reason guidelines shifted toward more permissive eGFR thresholds was growing evidence that metformin provides real cardiovascular and survival benefits, even for people with reduced kidney function. In the SAVOR-TIMI 53 trial, metformin users had a lower risk of death from any cause (roughly 25% lower, after statistical adjustment) compared to non-users, though the benefit did not reach statistical significance in the subgroup with moderate-to-severe kidney disease or heart failure.¹⁰

A large nationwide cohort study of patients with diabetes and chronic kidney disease found a more complex picture. Metformin users had a lower risk of death from any cause (about 22% lower) and a substantially lower risk of progressing to end-stage kidney disease (about 56% lower) compared to non-users. At the same time, major cardiovascular events were slightly more common in metformin users, a finding the authors suggested may reflect differences in baseline health between the groups.¹¹ Separate monitoring data support the position that metformin-treated patients with an eGFR above 30 have a reduction in mortality, reinforcing the case for continuing the drug down to that threshold rather than pulling it earlier.¹²

That said, continuing metformin is not clearly benign for the kidneys themselves in all patients. One study comparing patients who stayed on metformin versus those who stopped it found that the kidney function decline was faster in those who continued the drug. After adjusting for other factors, continuing metformin was associated with a steeper eGFR decline of roughly 2 mL/min/1.73 m² per year more than the group that stopped.¹³ This is a single observational study and should not be overweighted, but it underscores why the “caution zone” between eGFR 30 and 45 calls for closer monitoring and dose reduction rather than simply continuing at full strength.

Cystatin C and Getting a Better Kidney Estimate

Because creatinine is so heavily influenced by muscle mass, age, and sex, some clinicians are turning to cystatin C, a different protein filtered by the kidneys that is less affected by these variables. When researchers compared eGFR calculated from creatinine (eGFRcr) versus eGFR calculated from cystatin C (eGFRcys) in veterans with diabetes, the two measures disagreed substantially. About 20% of patients were reclassified into a different kidney function category by cystatin C, and the reclassifications were predominantly downward, meaning cystatin C suggested worse kidney function than creatinine did. Nearly 6% of patients were reclassified into a “do not use metformin” category, and about 13% were reclassified into a “use with caution” category.¹⁴

A separate study in elderly patients with type 2 diabetes reached a similar conclusion: cystatin C-based eGFR reclassified a significant proportion of patients from eligible to less-eligible or ineligible stages for metformin, recommending that both creatinine- and cystatin C-based eGFR be calculated for metformin dosing decisions in older adults.¹⁵ This matters most for older people and those with low muscle mass, precisely the populations where creatinine alone tends to paint an overly optimistic picture of kidney health. If you are elderly, thin, or have lost significant muscle, and your creatinine looks reassuringly normal, it may be worth asking whether a cystatin C test would change the picture.

Drugs That Interfere With Metformin Clearance

Even with stable kidney function, certain medications can change how your body handles metformin. The drug is transported into and out of kidney cells by specialized transporter proteins (primarily OCT2 and MATE transporters). Any medication that competes for those same transporters can slow metformin’s exit from the body, effectively raising its concentration in your blood.¹⁶ Common culprits include certain heart rhythm drugs and cimetidine (an older heartburn medication). Laboratory studies have shown that drugs like procainamide and quinidine can inhibit metformin transport through its kidney transporter even at normal therapeutic doses.¹⁷

This interaction becomes clinically relevant when someone is already near the eGFR threshold. If your kidneys are working at, say, 40% capacity, and a new medication further impairs metformin clearance, you could cross into unsafe territory without any change in your creatinine or eGFR. This is one reason why pharmacists and prescribers should review drug interactions whenever metformin is used alongside other medications, particularly in people with kidney function in the caution zone.

Can Metformin Actually Protect the Kidneys

One of the more interesting developments in metformin research is the suggestion that the drug may have protective effects on the kidneys independent of its blood sugar-lowering action. Animal studies of diabetic kidney disease have found that metformin reduces kidney damage markers, decreases tissue scarring, and protects the small filtering structures of the kidney. These effects appear to work through an anti-inflammatory pathway, reducing the activity of a signaling system involved in tissue damage and fibrosis.¹⁸ The large cohort study mentioned earlier, which found that metformin users had a dramatically lower risk of progressing to end-stage kidney disease, is at least consistent with this idea, though observational data can never prove causation.¹¹

The irony is hard to miss: a drug historically withheld from people with kidney disease over safety concerns may actually slow the progression of that kidney disease in certain patients. This does not mean anyone should take metformin with severely impaired kidneys. But it does explain why the medical community has steadily moved toward letting more patients with moderate kidney impairment stay on the drug, under close monitoring, rather than reflexively stopping it at an arbitrary creatinine number.

What to Ask Your Doctor

If you have been told your creatinine is elevated and you are wondering whether to stop metformin, the most useful question is not “what is my creatinine?” but “what is my eGFR?” Your doctor should already be calculating this. If your eGFR is above 45, you are generally in clear territory. If it is between 30 and 45, expect a dose reduction and more frequent lab checks. If it is below 30, metformin should be discontinued.

A few additional points worth raising with your provider: whether a cystatin C level might give a more accurate picture of your kidney function, especially if you are older or have low muscle mass; whether any of your other medications could interfere with metformin clearance; and whether you have a clear plan for what to do with metformin if you get acutely ill. Many patients are never told about sick-day rules, and a bad bout of food poisoning or a severe infection is exactly the scenario where a drug that was perfectly safe yesterday can become risky today.

1
JAMA Internal Medicine. Serum Creatinine Is an Inadequate Screening Test for Renal Failure in Elderly Patients
2
PubMed Central. Metformin Treatment for Patients with Diabetes and Chronic Kidney Disease: A Korean Diabetes Association and Korean Society of Nephrology Consensus Statement
3
PubMed. Optimizing metformin therapy in practice: Tailoring therapy in specific patient groups to improve tolerability, efficacy and outcomes
4
PubMed. Metformin induces lactate production in peripheral blood mononuclear cells and platelets through specific mitochondrial complex I inhibition
5
Toxicology Letters. Metformin-associated/induced lactic acidosis: an example of threatening drug-inced mitochondrial disease
6
JAMA Internal Medicine. Risk of Fatal and Nonfatal Lactic Acidosis With Metformin Use in Type 2 Diabetes Mellitus: Systematic Review and Meta-analysis
7
PubMed Central. Metformin exposure and the incidence of lactic acidosis in critically ill patients with T2DM: A retrospective cohort study
8
PubMed. Metformin associated lactic acidosis: incidence and clinical correlation with metformin serum concentration measurements
9
PubMed Central. Medication Holds in CKD During Acute Volume-Depleting Illnesses: A Randomized Controlled Trial of a “Sick-Day” Protocol
10
PubMed. Metformin Use and Clinical Outcomes Among Patients With Diabetes Mellitus With or Without Heart Failure or Kidney Dysfunction: Observations From the SAVOR-TIMI 53 Trial
11
PubMed Central. Metformin use and cardiovascular outcomes in patients with diabetes and chronic kidney disease: a nationwide cohort study
12
PubMed Central. Monitoring Patients on Metformin: Recent Changes and Rationales
13
PubMed Central. Effect of metformin on kidney function in patients with type 2 diabetes mellitus and moderate chronic kidney disease
14
PubMed Central. Use of cystatin C to inform metformin eligibility among adult veterans with diabetes
15
Basic & Clinical Pharmacology & Toxicology. Cystatin C measurement leads to lower metformin dosage in elderly type 2 diabetic patients
16
PubMed Central. Drug Interactions of Metformin Involving Drug Transporter Proteins
17
PubMed. Metformin transport by renal basolateral organic cation transporter hOCT2
18
PubMed Central. Metformin and Canagliflozin Are Equally Renoprotective in Diabetic Kidney Disease but Have No Synergistic Effect