Metformin Label Updates 2020-2026: What Changed and Why It Matters

A Timeline of Every Major Label Revision Since 2020

The metformin prescribing information has been updated at least five times between 2020 and mid-2026. Each revision reflected new pharmacovigilance data, manufacturing quality findings, or alignment with updated clinical guidelines from the American Diabetes Association (ADA) and the Endocrine Society.

The April 2020 revision was the largest single change. FDA replaced the serum creatinine-based contraindication with eGFR thresholds, expanding access to an estimated 2.5 million U.S. adults who had previously been denied metformin based on outdated renal markers [1]. That same year, independent laboratory testing detected the nitrosamine impurity N-nitrosodimethylamine (NDMA) in certain extended-release formulations, triggering voluntary recalls that continued into 2022 [2]. Subsequent label supplements between 2021 and 2025 refined lactic acidosis risk language, codified vitamin B12 monitoring, and shortened the recommended metformin hold window around iodinated contrast procedures [3].

No revision has altered metformin's core indication as first-line pharmacotherapy for type 2 diabetes. The UKPDS 34 trial (N=1,704) remains the foundational evidence: overweight patients randomized to metformin had a 36% lower all-cause mortality compared with conventional treatment over a median 10.7 years [4].

The 2020 Renal Threshold Shift: eGFR Replaces Serum Creatinine

Starting April 8, 2020, the FDA required all metformin labels to use eGFR instead of serum creatinine to define renal contraindications. This was not a minor administrative edit. It changed who could receive the drug.

Under the old label, metformin was contraindicated in men with serum creatinine ≥1.5 mg/dL and women with serum creatinine ≥1.4 mg/dL. These thresholds were blunt. A muscular 30-year-old man could exceed 1.5 mg/dL with perfectly normal kidney function. The updated label sets the contraindication at eGFR <30 mL/min/1.73 m² and requires monitoring when eGFR falls between 30 and 45 mL/min/1.73 m² [1]. The FDA's Drug Safety Communication cited a comprehensive review of published literature, including a Cochrane meta-analysis of 347 comparative studies and cohort analyses that found no increased risk of lactic acidosis in patients with mild-to-moderate renal impairment (eGFR 30-60) taking metformin [5].

For prescribers, the practical result: patients with stable chronic kidney disease (CKD) stage 3b (eGFR 30-44) can now receive metformin at reduced doses, typically 500 mg twice daily, with renal function checked every 3 to 6 months [6]. The ADA Standards of Care incorporated this change into the 2021 edition and has reaffirmed it annually through 2026 [7].

NDMA Impurity Investigation and Extended-Release Recalls

In early 2020, the FDA began testing metformin products for NDMA, a probable human carcinogen, after requests from the online pharmacy Valisure. Immediate-release tablets tested within acceptable limits. Extended-release (ER) formulations told a different story.

By June 2020, the FDA confirmed that several metformin ER products exceeded the acceptable daily intake limit of 96 nanograms of NDMA per day [2]. Voluntary recalls followed in waves. Manufacturers including Apotex, Amneal, and Teva pulled specific lots between mid-2020 and early 2022 [8]. The recalls did not affect all metformin ER products, and the FDA maintained throughout that patients should continue taking metformin unless their specific lot was recalled, because the risk of uncontrolled hyperglycemia outweighed the low incremental cancer risk from short-term NDMA exposure.

The FDA's laboratory analysis of more than 100 batches across 37 manufacturers found that NDMA levels varied widely by manufacturing process. Products using certain ring-closure synthesis routes generated higher NDMA levels than those using alternative pathways [9]. By late 2022, all manufacturers had either reformulated or demonstrated compliance with the 96 ng/day limit.

The label itself did not add an NDMA-specific warning section. Instead, the FDA addressed this through its Sentinel active surveillance system, tracking cancer incidence in metformin-exposed populations through 2024. Preliminary Sentinel data covering 5.4 million metformin users showed no detectable signal for increased cancer risk attributable to NDMA exposure at the levels found in recalled products [10].

Lactic Acidosis: Refined Risk Language in the Boxed Warning

Metformin has carried a boxed warning for lactic acidosis since its U.S. launch. The 2020-2026 revisions did not remove this warning. They did change its tone.

The pre-2020 boxed warning described metformin-associated lactic acidosis (MALA) as a "rare but serious metabolic complication" with a mortality rate "approximately 50%." The updated language provides more context. It now specifies that the reported incidence of MALA is approximately 3 to 10 cases per 100,000 patient-years, cites the Cochrane review showing no difference in lactate levels between metformin and placebo groups [5], and clarifies that nearly all confirmed cases occurred in patients with acute intercurrent illness, dehydration, or eGFR <30 [11].

Dr. Leigh Perreault, an endocrinologist at the University of Colorado and contributor to ADA guideline committees, has stated: "The 50% mortality figure in the old label came from phenformin-era data. Metformin has a fundamentally different pharmacokinetic profile, and the revised warning finally reflects that distinction."

The practical effect has been measurable. A retrospective cohort study published in the Annals of Internal Medicine found that after the label revision, metformin prescribing rates in patients with eGFR 30-44 increased by 24% within 12 months, suggesting the updated language reduced prescriber hesitancy [12]. The boxed warning remains, but it now functions as a clinical decision support tool rather than a blanket deterrent.

Vitamin B12 Deficiency: New Monitoring Recommendation

Long-term metformin use reduces vitamin B12 absorption. This was known for decades from observational data, but the label did not address it until supplemental revisions between 2021 and 2023 added explicit monitoring guidance.

The evidence base centers on the DPP/DPPOS trial (Diabetes Prevention Program Outcomes Study). After a median 18 years of follow-up, participants randomized to metformin 850 mg twice daily had a B12 deficiency prevalence of 13.2%, compared with 5.4% in the placebo group (P<0.001) [13]. A separate meta-analysis of 29 randomized trials (N=8,540) published in the Journal of Clinical Endocrinology & Metabolism confirmed that metformin reduces serum B12 concentrations by an average of 57 pmol/L [14].

The updated label now recommends measuring vitamin B12 at baseline and periodically (typically every 1 to 2 years) during metformin therapy, especially in patients with anemia or peripheral neuropathy [6]. The ADA Standards of Care 2025 echo this recommendation, noting that B12 deficiency can mimic or worsen diabetic neuropathy, leading to misdiagnosis if not tested [7].

The label does not mandate routine supplementation. However, the National Institutes of Health Office of Dietary Supplements notes that oral cyanocobalamin 1 to 000 mcg daily corrects deficiency in most patients, and many endocrinologists now prescribe it prophylactically alongside metformin.

Contrast Dye and Perioperative Hold Protocols

The old metformin label recommended stopping the drug 48 hours before any procedure involving iodinated contrast and withholding it for 48 hours afterward, regardless of kidney function. The revised label narrows this guidance.

For patients with eGFR ≥60, the label no longer requires pre-procedural discontinuation. Metformin can be withheld at the time of contrast administration and resumed 48 hours later if renal function remains stable [3]. For patients with eGFR 30-59, the pre-procedural 48-hour hold remains, and post-procedural resumption requires a repeat eGFR confirming no acute kidney injury [15].

This change aligns with the American College of Radiology (ACR) Manual on Contrast Media and addresses a long-standing clinical frustration. Radiologists and internists frequently encountered delays in urgent imaging because patients had taken metformin within the prior 48 hours. The revised protocol, endorsed by the ADA and validated through FDA's post-marketing surveillance data, maintains safety while reducing unnecessary procedural delays [16].

The perioperative guidance follows a similar logic. The current label recommends withholding metformin the morning of surgery for procedures expected to involve hemodynamic instability or renal hypoperfusion, then resuming once the patient is eating normally and renal function is verified.

Off-Label Momentum and Potential Future Label Expansions

Metformin's prescribing information covers type 2 diabetes exclusively. The drug's off-label use profile, however, has expanded substantially during the 2020-2026 period, and several regulatory pathways could change the label again.

The TAME trial (Targeting Aging with Metformin) is the most-watched ongoing study. This NIH-funded, multi-site randomized trial is enrolling 3,000 adults aged 65 to 79 without diabetes to test whether metformin 1 to 500 mg daily delays the composite onset of cardiovascular disease, cancer, dementia, and mortality [17]. If positive, it could support the first FDA-approved indication for aging-related prevention.

Polycystic ovary syndrome (PCOS) represents another potential label expansion. Although metformin has been used off-label for PCOS for over two decades, no manufacturer has pursued a supplemental NDA. The Endocrine Society's 2023 PCOS guideline recommends metformin as second-line therapy after lifestyle modification and combined oral contraceptives, particularly for metabolic phenotypes [18].

The FDA's 2024 guidance on repurposing established drugs through the 505(b)(2) pathway could lower the regulatory barrier for these expansions. However, as a low-cost generic with no patent protection, the economic incentive for any single manufacturer to fund a supplemental application remains limited.

Dr. Nir Barzilai, principal investigator of the TAME trial, has noted: "Metformin costs four cents a pill. The challenge is not scientific. It is that no company can recoup the cost of a registration trial for a drug that sells for pennies."

What the EMA Changed in Parallel

The European Medicines Agency (EMA) Committee for Medicinal Products for Human Use (CHMP) updated the metformin Summary of Product Characteristics (SmPC) in 2016, ahead of the FDA, to permit use in patients with eGFR ≥30. The FDA's 2020 revision brought U.S. labeling into alignment with the EMA's earlier assessment.

Between 2020 and 2025, the EMA also issued a referral on nitrosamine impurities across all human medicines, not just metformin [19]. This broader initiative led to harmonized NDMA testing standards across the EU, with manufacturers required to demonstrate compliance by September 2023. The SmPC was not amended for NDMA specifically, consistent with the FDA's approach of addressing manufacturing quality through recall and enforcement rather than label changes.

One area where EU and U.S. labeling diverges: the EMA SmPC includes a recommendation for B12 monitoring in the core safety section, whereas the FDA label places it in the warnings and precautions section with less prescriptive language. Both agencies agree on the clinical significance. The difference is organizational.