de Jager · BMJ (Clinical research ed.) 2010 · randomized placebo-controlled trial · n=390

Long term treatment with metformin in patients with type 2 diabetes and risk of vitamin B-12 deficiency: randomised placebo controlled trial.

Cited 693 times in the scientific literature.

Level 2 - randomized trial

Multicenter randomized placebo-controlled trial

PubMed 20488910 · doi:10.1136/bmj.c2181 · record verified 2026-08-30

What was done

A multicenter randomized placebo-controlled trial evaluated 390 patients with type 2 diabetes receiving insulin across outpatient clinics of three non-academic hospitals in the Netherlands. Participants were randomized to receive either 850 mg metformin or placebo three times daily for 4.3 years. Percentage changes from baseline in vitamin B-12, folate, and homocysteine concentrations were assessed at 4, 17, 30, 43, and 52 months, along with rates of vitamin B-12 deficiency (<150 pmol/l) and low B-12 status (150–220 pmol/l).

What was found

Compared with placebo, metformin was associated with a mean decrease in vitamin B-12 concentration of -19% (95% CI -24% to -14%; P<0.001), a decrease in folate of -5% (95% CI -10% to -0.4%; P=0.033, which became non-significant after adjustment for BMI and smoking), and an increase in homocysteine of 5% (95% CI -1% to 11%; P=0.091). At study end, the absolute risk of vitamin B-12 deficiency was 7.2 percentage points higher with metformin than placebo (95% CI 2.3 to 12.1; P=0.004; number needed to harm [NNH] 13.8 per 4.3 years). The absolute risk of low vitamin B-12 concentration was 11.2 percentage points higher (95% CI 4.6 to 17.9; P=0.001; NNH 8.9 per 4.3 years). Patients with B-12 deficiency had significantly higher mean homocysteine levels (23.7 micromol/l) than those with normal B-12 levels (14.9 micromol/l; P<0.001).

Why it matters

This trial confirms that long-term metformin directly induces progressive vitamin B-12 depletion and secondary homocysteine elevation. It supports routine monitoring of vitamin B-12 levels in patients on chronic metformin therapy.

Limits

The study was confined to insulin-treated type 2 diabetes patients in Dutch hospital outpatient clinics, which may limit generalizability to diet-controlled or non-insulin-dependent populations. The abstract does not report clinical endpoints such as neuropathy, cognitive changes, or anemia.

Cited by