Vitamin B12: Drugs That Distort This Test

What Vitamin B12 Means and Why Accuracy Matters

Vitamin B12 (cobalamin) is a water-soluble vitamin required for DNA synthesis, red blood cell formation, and myelin maintenance in the nervous system. A serum B12 test measures total circulating cobalamin bound to transcobalamin and haptocorrin transport proteins. The test matters because prolonged deficiency causes irreversible peripheral neuropathy, megaloblastic anemia, and cognitive decline [1].

The normal reference range at most U.S. laboratories falls between 200 and 900 pg/mL (148 to 664 pmol/L), though thresholds vary slightly across platforms [2]. Values below 200 pg/mL are generally considered deficient. The gray zone sits between 200 and 300 pg/mL, where roughly 30 percent of patients already have tissue-level depletion that serum testing alone misses. According to the National Institutes of Health Office of Dietary Supplements, "serum vitamin B12 levels may not accurately reflect intracellular concentrations" [3]. This disconnect is exactly where drug-induced distortion becomes clinically dangerous: a patient on metformin or nitrous oxide may have a serum value that reads as borderline-normal while functional deficiency is already damaging neurons.

Approximately 6 percent of U.S. adults under 60 and nearly 20 percent of those over 60 have B12 deficiency or marginal status [3]. When medications push results into false-normal territory or mask true depletion, clinicians miss diagnoses. The eight drug classes covered below each distort B12 testing through distinct mechanisms, requiring different clinical responses.

Metformin: The Most Common Offender

Metformin reduces serum B12 by impairing calcium-dependent absorption in the terminal ileum. This is a true depletion, not just an assay artifact. The American Diabetes Association (ADA) recommends periodic B12 measurement in patients on long-term metformin, especially those with anemia or neuropathy [4].

The evidence is substantial. A randomized, placebo-controlled trial by de Jager et al. (N=390) published in the BMJ found that metformin 850 mg three times daily reduced serum B12 by 19 percent over 4.3 years, with 7.2 percent of metformin users developing B12 deficiency versus 2.3 percent in the placebo group [5]. A cross-sectional analysis from the National Health and Nutrition Examination Survey (NHANES) reported that metformin users had B12 levels approximately 50 pg/mL lower than matched non-users [6]. The DPP/DPPOS (Diabetes Prevention Program Outcomes Study) follow-up showed that after 13 years of metformin use, the odds of B12 deficiency (serum B12 <200 pg/mL) were roughly double compared to placebo [7].

The practical implication is clear. Patients taking metformin for more than 4 years should have serum B12 checked annually. If values fall below 300 pg/mL, order MMA to confirm functional status before attributing neuropathy symptoms to diabetic neuropathy alone. B12 supplementation (1 to 000 mcg oral daily) typically corrects the deficit without requiring metformin discontinuation [4].

Proton Pump Inhibitors and H2 Blockers

PPIs (omeprazole, esomeprazole, lansoprazole, pantoprazole) and H2 receptor antagonists (famotidine, ranitidine) suppress gastric acid production. Gastric acid is required to release protein-bound B12 from food so that intrinsic factor can bind free cobalamin in the duodenum. Chronic acid suppression disrupts this first step.

A large nested case-control study by Lam et al. (N=25,956 cases) published in JAMA found that 2 or more years of PPI use was associated with a 65 percent increased risk of B12 deficiency (OR 1.65 to 95% CI 1.58 to 1.73) [8]. H2 blockers carried a smaller but still significant risk (OR 1.25 to 95% CI 1.17 to 1.34) with similar duration thresholds. The risk was dose-dependent: patients taking more than 1.5 PPI pills per day had higher odds of deficiency than those on lower doses [8].

The American Gastroenterological Association's 2017 clinical practice update noted that "patients on long-term PPI therapy should be monitored for potential adverse effects including vitamin B12 deficiency" [9]. A key clinical nuance separates PPI-induced B12 depletion from pernicious anemia. PPIs impair food-bound B12 absorption but do not block absorption of crystalline (supplemental) B12, because free-form B12 does not require acid-mediated protein separation. Patients who test low on PPIs can supplement with oral crystalline B12 (500 to 1 to 000 mcg daily) and typically restore normal levels within 2 to 3 months.

Screening should begin after 2 years of continuous PPI therapy. Obtain serum B12 and, if borderline (200 to 300 pg/mL), follow up with MMA.

Oral Contraceptives and Estrogen-Containing Therapies

Estrogen-containing oral contraceptives have been associated with lower serum B12 since the 1960s, but the mechanism remains debated. The leading hypothesis involves estrogen-induced redistribution of B12 from haptocorrin (a transport protein) into tissues, reducing circulating levels without causing true deficiency. A separate possibility is reduced intestinal absorption through altered bile acid metabolism.

A systematic review by Green et al. in the American Journal of Clinical Nutrition found that up to 19 percent of oral contraceptive users had serum B12 below 200 pg/mL, compared to 5 percent of matched controls [10]. The clinical significance depends on functional markers. In most OC users with low serum B12, MMA and homocysteine remain normal, suggesting a distributional shift rather than true tissue depletion. However, women who combine oral contraceptives with other risk factors (vegetarian diet, concurrent PPI use, prior gastric surgery) can develop genuine deficiency.

For women on combined oral contraceptives, checking B12 alone can be misleading. If serum B12 falls below 300 pg/mL, MMA testing distinguishes real deficiency from estrogen-mediated redistribution. Clinicians should not reflexively supplement based on serum B12 alone in this population.

Nitrous Oxide: Normal Levels, Inactive Enzyme

Nitrous oxide (N2O) irreversibly oxidizes the cobalt atom in methylcobalamin, converting active B12 into an inactive form. This destroys function without changing the total amount of circulating cobalamin. The standard serum B12 immunoassay cannot differentiate between active and inactivated B12 molecules, making it completely blind to N2O-induced deficiency [11].

This problem presents in two populations. Surgical patients receiving prolonged N2O anesthesia (particularly those with pre-existing borderline B12 status) may develop acute megaloblastic changes within days. Recreational N2O users ("whippets") may present with subacute combined degeneration of the spinal cord while serum B12 reads within normal range [11].

A case series published in the Journal of Clinical Neuroscience documented 27 patients with N2O-related myeloneuropathy. Mean serum B12 was 312 pg/mL, well within the normal reference range, yet all patients had markedly elevated MMA (mean 1,450 nmol/L; normal <370 nmol/L) [12]. The disconnect is stark.

The only reliable screen in N2O-exposed patients is MMA paired with homocysteine. Serum B12 should not be used in isolation for anyone with known or suspected N2O exposure. Treatment requires high-dose intramuscular B12 (1 to 000 mcg daily for 7 days, then weekly) and complete N2O cessation.

Anticonvulsants and Antimetabolites

Several anticonvulsants reduce B12 levels through interference with folate metabolism, impaired absorption, or increased hepatic clearance. Phenytoin, phenobarbital, carbamazepine, and valproic acid have all been associated with lower serum B12 in observational studies, though effect sizes vary [13].

A study in Epilepsia (N=226 patients on long-term anticonvulsant monotherapy) found that 18 percent had serum B12 below 200 pg/mL, compared to 4 percent in age-matched controls [13]. Phenytoin users showed the largest reductions. The mechanism appears multifactorial: phenytoin inhibits folate-dependent enzymes that recycle B12 through the methionine synthase pathway, and some anticonvulsants independently impair intestinal absorption.

Methotrexate operates differently. As a dihydrofolate reductase inhibitor, methotrexate blocks the methylation cycle that requires B12 as a cofactor. Serum B12 levels may remain normal, but functional B12 activity drops because the downstream pathway is pharmacologically disabled [14]. MMA may or may not rise depending on which metabolic pathway is most affected. Clinicians managing patients on methotrexate for rheumatoid arthritis, psoriasis, or oncologic indications should monitor both B12 and folate, and many protocols mandate concurrent folic acid 1 mg daily to mitigate hematologic toxicity.

Colchicine, used for gout and familial Mediterranean fever, reduces B12 absorption by disrupting ileal mucosal cell function at high doses or with prolonged use [15]. The effect is typically modest and reversible upon dose reduction.

Biotin and Other Supplement Interference

High-dose biotin creates a different problem entirely. Rather than depleting B12, biotin interferes with streptavidin-biotin-based immunoassays used by many laboratory platforms. Biotin doses exceeding 5 mg per day (common in hair and nail supplements, and prescribed at 100 to 300 mg daily for multiple sclerosis) can produce falsely normal or falsely elevated B12 results depending on the assay architecture [16].

The FDA issued a safety communication in 2017 warning that "biotin in patient samples can cause significantly incorrect results on laboratory tests that use biotin technology" [17]. Competitive immunoassays (used for some B12 platforms) may report falsely elevated values, while sandwich immunoassays may report falsely low values. The direction of error depends on the specific analyzer.

The fix is simple. Patients should discontinue biotin supplements for at least 72 hours before B12 testing. Laboratories are increasingly adding biotin interference warnings to requisition forms, but clinicians should ask about supplement use directly. This is a pre-analytical variable that no confirmatory test can correct after the sample is drawn.

Ascorbic acid (vitamin C) in very high doses (>1 to 000 mg taken shortly before a blood draw) has also been reported to interfere with some B12 assays, though this effect is less well documented and less clinically significant than biotin interference [16].

How to Confirm True B12 Status When Medications Interfere

When any drug on the list above is present, serum B12 alone is insufficient. A two-tier confirmation strategy is the clinical standard.

First, order methylmalonic acid (MMA). MMA rises when B12-dependent methylmalonyl-CoA mutase activity drops. An MMA above 370 nmol/L in the setting of low-normal B12 (200 to 400 pg/mL) confirms functional deficiency with high specificity [18]. MMA is the single most useful confirmatory marker because it is specific to B12 (unlike homocysteine, which also rises with folate deficiency, hypothyroidism, and renal impairment).

Second, check homocysteine. Elevated homocysteine (>13 micromol/L) supports B12 deficiency but is less specific. The combination of elevated MMA and elevated homocysteine in a patient on a B12-distorting medication makes the diagnosis near-certain.

The Endocrine Society and AACE clinical practice guidelines recommend using MMA as the primary confirmatory test when serum B12 falls between 200 and 400 pg/mL, regardless of medication status [19]. For patients on metformin or PPIs, the ADA recommends annual serum B12 with reflex MMA testing if B12 is below 300 pg/mL [4].

Holotranscobalamin (holoTC), also called "active B12," measures the biologically available fraction bound to transcobalamin II. It responds earlier to B12 depletion than total serum B12 and is less affected by estrogen-mediated redistribution [20]. HoloTC assays are increasingly available at reference laboratories and may eventually replace total B12 as the first-line screen, particularly for patients on medications that distort total B12.

How to Raise or Lower Vitamin B12 Safely on Medication

Raising B12 when medications cause depletion follows a predictable hierarchy. Oral crystalline cyanocobalamin at 1 to 000 mcg daily corrects most PPI-induced and metformin-induced deficiency within 2 to 3 months because crystalline B12 bypasses the acid-dependent and intrinsic-factor-dependent steps at pharmacologic doses [3]. Sublingual methylcobalamin (1 to 000 mcg) is an alternative with similar bioavailability. Intramuscular cyanocobalamin (1 to 000 mcg weekly for 4 weeks, then monthly) is reserved for patients with severe deficiency (serum B12 <100 pg/mL), neurologic symptoms, or absorption issues that oral dosing cannot overcome [1].

As the ADA Standards of Medical Care state, "consideration should be given to periodic monitoring of vitamin B12 levels in metformin-treated patients, especially in those with anemia or peripheral neuropathy" [4].

Lowering B12 is rarely a clinical goal. Elevated B12 (>900 pg/mL) without supplementation can indicate hepatocellular disease, myeloproliferative disorders, or chronic kidney disease, and the priority is diagnosing the underlying cause rather than reducing the vitamin itself [2]. Discontinuing unnecessary B12 supplements normalizes levels within weeks if no underlying pathology exists.

For patients requiring ongoing treatment with a B12-distorting medication, the strategy is monitor-and-replace rather than discontinue the offending drug. Metformin's glycemic benefits, a PPI's control of erosive esophagitis, and anticonvulsant seizure prophylaxis all outweigh the manageable risk of B12 depletion when appropriate monitoring is in place. Annual serum B12 with reflex MMA testing costs approximately $50 to $150 without insurance and takes fewer than 48 hours to result at most commercial laboratories.