What Labs Should You Monitor While Taking Metformin?

Kidney function and vitamin B12 top the list of labs your doctor should be tracking while you take metformin, but the full picture extends well beyond those two. Metformin is one of the most prescribed medications in the world, and its safety profile is genuinely strong, yet it interacts with several body systems in ways that only show up on bloodwork. Some of these checks are universally recommended, others are situation-dependent, and a few are surprisingly under-discussed given the evidence behind them.

Kidney Function Is the Non-Negotiable Check

Your kidneys clear metformin from the body almost entirely unchanged. When kidney function drops, metformin accumulates, and at high enough levels the drug can contribute to a dangerous buildup of lactic acid in the blood. That makes estimated glomerular filtration rate (eGFR) the single most important lab to monitor. A joint consensus statement from Korean diabetes and nephrology societies laid out the thresholds clearly: metformin can be used safely when eGFR is at or above 45, should be limited to no more than 1,000 mg daily if eGFR falls between 30 and 44, and is contraindicated below 30.¹ These cutoffs are broadly consistent with guidelines from the American Diabetes Association and the FDA’s revised labeling.

Most prescribers check eGFR before starting metformin and then recheck it at least once a year. If you have any risk factors for kidney decline, like high blood pressure, older age, or a history of acute kidney injury, your doctor may check it more often. A systematic review focused on older adults recommended even considering discontinuation when the eGFR drops to 60 or below in people over 80 or those with gastrointestinal complaints.² The kidney-function check also matters before any procedure involving contrast dye, since those agents temporarily stress the kidneys and could push metformin levels into a risky range.³

Vitamin B12 Deserves Routine Screening

Metformin reduces your body’s ability to absorb vitamin B12 from the gut. This is not a rare side effect. A large real-world database study found that each additional year of metformin use raised the likelihood of B12 deficiency by about 5%, and people who had been on the drug for four years or more had roughly 41% higher odds of deficiency compared to shorter-term users.⁴ The Diabetes Prevention Program Outcomes Study, which followed participants for years after a major clinical trial, confirmed that long-term metformin use was associated with biochemical B12 deficiency and anemia, and recommended routine B12 testing for people on the drug.⁵

What makes this sneaky is that B12 deficiency develops slowly. You can be deficient for years before symptoms like numbness, tingling, fatigue, or cognitive changes become obvious. In someone with diabetes, those neurological symptoms can easily be mistaken for diabetic neuropathy, which means the actual cause goes untreated. A randomized trial lasting over four years showed that metformin not only lowered serum B12 but also progressively raised methylmalonic acid (MMA), a more sensitive marker of functional B12 deficiency. The rise in MMA correlated with worsening neuropathy scores.⁶ If your B12 level comes back borderline or low-normal and you have symptoms, asking for an MMA level can help clarify whether you actually have a tissue-level deficiency.

There is no universal consensus on how often to check B12, but many endocrinologists recommend a baseline level before or shortly after starting metformin, and then every one to two years thereafter, especially after the four-year mark when odds of deficiency climb steeply.

Complete Blood Count and What It Reveals

A complete blood count (CBC) is a standard lab that most people with diabetes get periodically anyway, but it carries extra information for metformin users. One of the hallmarks of B12 deficiency is macrocytosis, meaning your red blood cells grow larger than they should. A cross-sectional study in patients with type 2 diabetes on metformin found that prolonged treatment was associated with macrocytosis alongside low B12 and folate levels.⁷ A separate prospective study confirmed a significant negative correlation between B12 levels and mean corpuscular volume (the measure of red blood cell size), along with differences in white blood cell counts between metformin-treated and metformin-naive patients.⁸

In practical terms, if your mean corpuscular volume starts creeping upward on routine bloodwork, that is a flag to check your B12. It is a cheap, early clue that metformin may be depleting your stores before you ever notice symptoms.

Lactic Acid Is Rarely Checked Routinely, and That Is Usually Fine

Lactic acidosis is the complication most closely associated with metformin in the public imagination, but it is genuinely rare. A narrative review of the literature found that while metformin can raise lactate levels in users, most studies showed those levels staying within the normal range.⁹ When lactic acidosis does occur, it is almost always in the setting of something else going wrong: kidney failure, severe dehydration, liver disease, sepsis, or a major overdose. A study of 49 metformin-treated patients who developed lactic acidosis (defined as arterial lactate at or above 5 mmol/L with a blood pH at or below 7.35) found these were patients in acute crisis, not people on routine therapy.¹⁰

So while lactate levels are not part of routine monitoring for someone on stable metformin therapy, they matter in acute-care settings. If you end up in the emergency department with vomiting, severe dehydration, or signs of sepsis, your medical team should be checking a lactate and a basic metabolic panel. A cross-sectional analysis of U.S. diabetic adults with chronic kidney disease used serum bicarbonate as a proxy marker for acid-base status and found it informative for assessing metabolic acidosis risk in metformin users.¹¹ Your standard metabolic panel already includes bicarbonate, so any downward trend is visible without ordering anything extra.

Liver Function Tests Tell a More Nuanced Story

Metformin was long restricted in people with liver disease out of concern it could worsen hepatic function, but the evidence has moved firmly in the other direction. A meta-analysis of randomized controlled trials in people with non-alcoholic fatty liver disease found that metformin modestly but significantly reduced AST levels and was not associated with harm to the liver.¹² A randomized double-blinded trial comparing metformin and pioglitazone in fatty liver patients similarly concluded that metformin at 1 g per day was safe and had favorable effects on liver function tests.¹³

The reason liver enzymes still show up on your monitoring panel is not because metformin damages the liver. It is because the liver plays a central role in clearing lactate, and if liver function is severely impaired, the risk of lactic acidosis goes up regardless of the drug. Additionally, people with diabetes have high rates of fatty liver disease, and tracking ALT and AST over time gives your doctor a window into whether that condition is progressing. Think of liver labs on metformin as “monitoring the landscape” rather than “watching for drug toxicity.”

Magnesium Is the Most Overlooked Check

Low magnesium is common in people with diabetes, and metformin appears to make it worse. A study of patients in the Fremantle Diabetes Study found that those on metformin, whether alone or combined with a sulfonylurea, had significantly lower serum magnesium concentrations and a higher percentage of hypomagnesemia compared with patients managed on diet alone.¹⁴ A study of Puerto Rican older adults found that metformin use was associated with nearly four-fold higher odds of hypomagnesemia.¹⁵

The mechanism may involve both gastrointestinal losses (metformin causes diarrhea in many users, which wastes magnesium) and changes to kidney handling of the mineral. A case report and narrative review found that chronic metformin use suppresses a specific magnesium transport channel in the kidneys, though it may simultaneously reduce urinary magnesium excretion, making the net effect complex.¹⁶ What matters practically is that low magnesium can worsen insulin resistance, cause muscle cramps, contribute to heart rhythm problems, and amplify the effects of low potassium and calcium. Yet magnesium is not part of a standard metabolic panel. You have to specifically request it. If you take metformin and have symptoms like persistent cramps, fatigue, or irregular heartbeat, a serum magnesium level is worth asking about.

Thyroid Function and Metformin’s Subtle Effect on TSH

This one catches people off guard. A meta-analysis in the Journal of Clinical Endocrinology & Metabolism found that metformin lowered TSH levels in both overt hypothyroidism (patients already on thyroid replacement) and subclinical hypothyroidism (patients not on thyroid medication).¹⁷ A separate study found that metformin lowered TSH in people with type 2 diabetes and hypothyroidism, and even in euthyroid patients who had positive thyroid antibodies.¹⁸

However, not every study agrees. A randomized controlled trial of 46 patients with subclinical hypothyroidism found no significant change in TSH, free T4, or free T3 after six months of metformin compared to controls.¹⁹ The conflicting results suggest the TSH-lowering effect may depend on the underlying thyroid condition and possibly on dose and duration of metformin use. The practical takeaway: if you take both metformin and levothyroxine, be aware that your TSH might look artificially lower than it would without metformin. If your doctor adjusts your thyroid dose based solely on a dropping TSH, the metformin effect should be part of that conversation.

Blood Glucose and the Rare Hypoglycemia Question

Metformin is famously unlikely to cause low blood sugar on its own, which is one of its great advantages over sulfonylureas and insulin. But “unlikely” is not “impossible.” A case report documented a teenager who ingested a massive overdose of metformin (75 grams) and developed severe, recurrent hypoglycemia with blood sugars dropping as low as 15 mg/dL, despite serum insulin levels that were actually suppressed rather than elevated.²⁰ That scenario is obviously extreme and not relevant to people taking prescribed doses. Still, if you combine metformin with insulin or a sulfonylurea, the combination does raise hypoglycemia risk, and your doctor may ask you to check fasting glucose or wear a continuous monitor more frequently during dose adjustments.

For most people on metformin alone, the standard monitoring is hemoglobin A1c every three to six months, which reflects average blood sugar over the prior two to three months. Fasting glucose or post-meal checks are usually reserved for people adding other agents or those not reaching target A1c.

A Lab Interference Worth Knowing About

Here is something that rarely comes up in patient counseling but has real implications: metformin can interfere with certain lab assays. A recent study found that metformin causes a spurious decrease in urinary creatinine when measured by enzymatic methods. Since the urine albumin-to-creatinine ratio (UACR) depends on accurate creatinine measurement, falsely low creatinine from metformin interference can produce falsely elevated UACR values.²¹ UACR is a critical test for detecting early diabetic kidney disease, and a falsely high result could trigger unnecessary concern, additional testing, or even changes in treatment. If your UACR comes back unexpectedly high, it is worth discussing with your doctor whether the enzymatic creatinine assay used in your lab could have been affected by your metformin.

Therapeutic Drug Monitoring Is Not Standard Practice

Unlike many medications, metformin does not have a well-established blood-level range that clinicians routinely target. A systematic review in Clinical Pharmacokinetics pointed out that despite over 50 years of use, major methodological errors have confused the literature on what metformin’s therapeutic concentration window actually is. The only valid way to define it would be to correlate blood glucose control with plasma metformin concentration in long-term treated patients, and that work has not been done convincingly.²² A forensic toxicology study attempted to establish reference values and found an average plasma metformin concentration of about 1,846 ng/mL in clinical patients, with no correlation between plasma concentration and the prescribed daily dose in most dose groups.²³

In practice, this means your doctor gauges whether metformin is working by looking at your A1c and fasting glucose, not by measuring the drug itself. Metformin blood levels are essentially reserved for emergency departments dealing with overdoses or poisonings, and for forensic investigations.

GDF15 and the Frontier of Metformin Research

One emerging area involves growth differentiation factor 15, or GDF15, a hormone-like protein that metformin activates. A review found that metformin increases circulating GDF15 levels, which suppresses appetite and contributes to the modest weight loss many users experience.²⁴ GDF15 is not currently measured in clinical practice for metformin monitoring, but it is gaining interest as a biomarker that might explain why some people lose weight on metformin and others do not, and why gastrointestinal side effects are more severe in some users (the same GDF15 pathway that suppresses appetite can also trigger nausea).

Why Some People Respond Differently to Metformin

Genetic variation in a transporter protein called OCT1 helps explain why metformin works well for some people and barely works for others. OCT1 is the main channel that carries metformin into liver cells, where the drug does most of its glucose-lowering work. A landmark study in the Journal of Clinical Investigation found that people carrying reduced-function versions of the OCT1 gene had significantly weaker responses to metformin on glucose tolerance tests.²⁵ Certain OCT1 variants have also been linked to metformin intolerance and gastrointestinal side effects.²⁶

Pharmacogenomic testing for OCT1 variants is not standard before prescribing metformin, but a meta-analysis of available studies confirmed that OCT1 polymorphisms influence measures like fasting glucose and A1c response to the drug.²⁷ If you have tried metformin at adequate doses and your glucose levels barely budged, or if the side effects are intolerable even at low doses, this genetic variability is a plausible explanation. Some academic medical centers now offer pharmacogenomic panels that include OCT1, though this remains far from routine.

Putting a Monitoring Schedule Together

Not every lab on this list needs to be checked at the same frequency, and not every lab applies to every patient. As a rough framework:

Before starting: eGFR, baseline B12, liver enzymes, A1c, and a basic metabolic panel. Thyroid function if there is any history of thyroid disease.
Every 3-6 months: A1c. Fasting glucose if you are also on insulin or a sulfonylurea.
Annually: eGFR, basic metabolic panel (which includes bicarbonate and electrolytes), liver enzymes. Consider UACR for diabetic kidney screening, with awareness of the assay interference issue.
Every 1-2 years: Vitamin B12, especially after year four. Add MMA if B12 is borderline and you have neurological symptoms.
As needed: Magnesium if you have symptoms of deficiency. TSH if you take thyroid medication and your levels seem to be shifting without explanation. Lactate only in acute illness or emergency settings.

Individual circumstances change the schedule. If your eGFR is trending downward, it may need checking every three to six months rather than annually. If you are over 80, monitoring generally tightens across the board. If you develop new gastrointestinal symptoms after years of tolerating metformin, B12 and magnesium are both reasonable things to check before assuming the drug has simply stopped agreeing with you.