How do you approach the use of metformin in the management of prediabetes for patients who are neither obese nor have a family history of diabetes?

When considering the approach to a patient with a diagnosis of pre-diabetes, I think it is important to be sure the patient actually meets criteria. Most of us screen patients at risk with A1C levels because it is so much easier. However, A1C levels do not fully correspond to glucose levels and may underestimate the actual occurrence of diabetes. So for patients with A1C close to the diabetic range, I think it would be worthwhile to check an OGTT if at all possible to see if the patient meets criteria for diabetes, because the management algorithm will change. Similarly, for normal-weight patients with A1C in the lowest range, checking a fasting glucose to confirm the pre-diabetes state would be helpful.

Once everyone agrees with the diagnosis of pre-diabetes, then have a discussion of management with the patient. I think we all turn to the results of the Diabetes Prevention Program to help counsel our patients. In that population of very high-risk patients (basically impaired fasting glucose and impaired glucose tolerance) diet and exercise slowed the development of diabetes by nearly 60% and at best metformin slowed it by about 30%, so metformin was far less effective. When breaking it down by subgroups metformin was essentially ineffective in the non-obese patients and the older patients. I have a subgroup analysis below:

I think this shows how powerful lifestyle changes were and metformin only comes close in small specific groups of patients. Metformin isn’t really better in any group. I encourage lifestyle to the extent the patient is able to comply and reserve offering metformin to the very obese and younger patients.

Metformin is not approved by the FDA for use in prediabetes. NO drug is approved for the treatment of Prediabetes. Thus, their use would be considered off-label. However, the American diabetes association states that the use of Metformin can be considered in subjects with a BMI of 35 or more, patients younger than 60, and women with gestational diabetes.