It is challenging to know the extent to which metformin might be contributing to the neuropathy that we see in patients with type 2 diabetes for a number of reasons, including the fact that metformin is so commonly prescribed and the variable manifestations of diabetic neuropathy. For instance, the most common form of diabetic neuropathy, distal symmetric polyneuropathy, is already present to some degree in at least 10% to 15% of patients newly diagnosed with type 2 diabetes. Further, the rates of diabetic neuropathy vary considerably depending on the test used (eg, nerve conduction studies will identify neuropathy in many more patients than we can detect clinically).

If we accept the assumption that neuropathy is primarily caused by high blood sugar, which is supported by the Diabetes Control and Complications Trial, then a key question becomes: Could metformin be responsible for accelerating the progression of diabetic neuropathy? Based on the available evidence, the most likely mechanism involves vitamin B₁₂ and perhaps even folic acid, since the absorption of both seems to be impaired with metformin. We do see more vitamin B₁₂ deficiency in long-term metformin users in the Diabetes Prevention Program Outcomes Study cohort. At 5 years, a low/borderline low vitamin B₁₂ levels were more common in the metformin group than in the placebo group (19.1% vs 9.5%), with the likelihood of low vitamin B₁₂ increasing over time. 

Vitamin B₁₂ deficiency is associated with peripheral neuropathy and loss of sensation in peripheral nerves. Such deficiency may cause damage to specific nerves in the nerve bundle that are also commonly involved in neuropathy due to high blood sugar (ie, large-fiber nerves). Patients with neuropathy from vitamin B₁₂ deficiency may have symptoms of imbalance, which are also common in diabetic neuropathy.

Given the evidence, I would suggest that we should be screening for vitamin B₁₂ deficiency every year in patients who have been on metformin for 5 years or longer. The test is relatively inexpensive, and vitamin B₁₂ deficiency is associated with many conditions that overlap with diabetic neuropathy, including pain, risk of falls, ulcerations, and amputations. I suspect that our guidelines will likely evolve to include additional language regarding the need for such screening and the best methods for detection, although the inherent challenges in such recommendations include consensus on the best screening test(s) to use (eg, vitamin level alone or with methylmalonic acid). Many patients ask if they should avoid using metformin given the risk of vitamin deficiency, especially since metformin does not have the major impact on cardiovascular outcomes that is observed with some of the other agents in our toolbox. But since glucose control is primary in the prevention of diabetic neuropathy and metformin overall has an acceptable side-effect profile, if a patient is achieving glycemic targets while on metformin, I would emphasize detection and treatment of the vitamin deficiency rather than switching to another medication. Also important to consider is that since vitamin B₁₂ deficiency is relatively common, especially in the older adult, metformin is not always the clear culprit. There are many other potential causes that should be explored when deficiency is diagnosed. However, if patients raise concerns regarding metformin, I support them in treating their diabetes with non-metformin options, as long as they can do so and maintain good glycemic control.