STOCKHOLM – Clinicians might want to consider more routinely monitoring levels of vitamin B12 in their diabetic patients who are using metformin, according to new data from the HOME (Hyperinsulinaemia: the Outcome of Its Metabolic Effects) randomized controlled trial, which showed that B12 depletion was linked to neurotoxicity.
The new study findings, presented at the annual meeting of the European Association for the Study of Diabetes, showed that metformin increased levels of serum methylmalonic acid (MMA), the gold standard biomarker of tissue B12 deficiency, and that this in turn was linked to neuropathy measured using a validated score.
“We earlier showed that metformin causes B12 deficiency (<150 pmol/L) with a number needed to harm of 14 after 4.3 years,” said study investigator Dr. Mattijs Out of Bethesda Diabetes Research Center in Hoogeveen, the Netherlands, referring to the original results of the trial (BMJ. 2010;340:c2181).
“Today I showed you that metformin increases MMA, dose dependently and progressively over time,” he added. “Metformin has two opposite effects on neuropathy,” he continued, “a neuroprotective effect by improving glycemic control and a neurotoxic effect by inducing B12 depletion.”
Dr. Out noted that metformin remains the cornerstone of type 2 diabetes therapy with over 100 million prescriptions per year, putting many patients at risk for B12 deficiency and its consequences.
Current guidelines from the American Diabetes Association and EASD mention B12 deficiency as a potential downside of metformin use but do not go so far as giving specific recommendations because of lack of evidence on whether screening should be routinely performed or if vitamin B12 supplements should be given.
“Consequences of vitamin B12 deficiency, such as neuropathy or mental changes, may be profound,” Dr. Out said. Changes because of B12 deficits can be difficult to diagnose because they may be ascribed to old age or diabetes itself, and they may become irreversible if unchecked, he observed. Diagnosing and managing B12 deficiency is, in theory, “easy, cheap, and effective,” he said. A new trial would be needed, however, to determine whether screening for B12 deficiency or supplementation would be the better approach.
In the current study, structural equation monitoring (SEM) was used to determine the likely effects of metformin on the Valk neuropathy score (Diabetes Med. 1992;9:716-21) directly and via effects on hemoglobin A1c and MMA. The analysis involved 390 insulin-treated patients with type 2 diabetes who were also treated with 850 mg metformin or placebo up to three times daily for 52 months. Over a 4.3-year follow-up period, patients had HbA1c and neuropathy scores measured 17 times, and MMA was measured at six visits.
Compared with placebo, metformin use was associated with a significant (P= .001) 0.039 micromol/L (95% confidence interval, 0.019-0.055) increase in MMA over the course of the study.
While there was no significant difference in neuropathy scores between the placebo- and metformin-treated groups, SEM showed that metformin had a beneficial effect on lowering the neuropathy score via lowering HbA1c and an adverse effect on the neuropathy score by increasing MMA levels. Overall, metformin was associated with a 0.25-point increase in the neuropathy score, suggesting the net effect of the oral hypoglycemic drug is a negative one as higher scores mean worse neuropathy.
After the presentation of the findings, session chair Dr. Guntram Schernthaner commented: “In my center in Vienna, we have been using metformin for 30 years, but we have never used this high dosage [850 mg]. We have suspected several times the B12 deficiency and we measured it in many cases, but we never found any cases of severe B12 deficiency.”
Dr. Schernthaner, who is professor of medicine at the University of Vienna in Austria, suggested the next step would be to perform a large trial of maybe 2,000 metformin users to see how often B12 deficiency really occurs.
Dr. Out agreed a larger trial would be the ideal and conceded that the effect size was small. However, with such a large number of patients using the drug worldwide, potentially “many patients are at risk.”
He said a direct effect of metformin on neuropathy would perhaps not be found by general screening because of its protective effect via improved glycemic control, “so you may miss B12 deficiency–induced neuropathy.” Measuring MMA, where possible, may be a solution.