Study Finds Metformin Works Better in African Americans

Miriam E. Tucker

June 12, 2014

African Americans with type 2 diabetes respond better to metformin's glucose-lowering effect than do whites, a new study finds.

The results, from a large health system electronic records database, were published online June 12 in the Journal of Clinical Endocrinology & Metabolism.

"We were quite surprised by the consistency of our findings," said lead author L. Keoki Williams, MD, associate director of the Center for Health Policy and Health Services Research at the Henry Ford Health System in Detroit, Michigan. "Regardless of how well-controlled diabetes was at baseline, African American patients appeared to respond consistently better to metformin than did white patients," he told Medscape Medical News.

The findings lend support to current recommendations to use metformin as first-line treatment for type 2 diabetes in this population in particular, he said.

However, he added, "Further studies are needed to assess whether the greater reduction in HbA1c levels in African Americans also translates to greater improvement in hard clinical outcomes."

Consistent Findings

Dr. Williams and colleagues identified 19,672 patients with diabetes who had records of 2 or more prescription fills of metformin between January 1997 and June 2013 and who had at least 2 HbA1c measurements separated by 4 months between the first and last metformin prescriptions on record. Of those individuals, 7429 self-identified as being African American, and 8783 as white.

Patients in the African American group were younger (55.1 vs 59.5 years; P = .001), more were female (58.6% vs 47.8%; P = .001), and they had higher baseline HbA1c values (7.8% vs 7.4%; P = .001).

The investigators considered the first HbA1c measurement after metformin initiation as the baseline value so that changes in HbA1c levels could be related to the measures of drug exposure in the intervening time period while adjusting for total time on treatment. The average duration of time between metformin initiation and baseline HbA1c measurement was 171 days, and average follow-up was 5.18 years.

After adjustment for age, sex, race/ethnicity, body mass index, baseline HbA1c, total time on metformin, and exposure to other diabetes medications, metformin use at the maximum dose (2550 mg/day) was associated with an overall 0.62-percentage-point reduction in HbA1c compared with no metformin exposure.

The results showed a large difference by race/ethnicity. Metformin use was associated with a 0.90-percentage-point HbA1c reduction from baseline among African Americans compared with a drop of just 0.42-percentage-points among whites (P = .001). The differences between the 2 groups remained significant for all baseline HbA1c values between 6% and 9%.

"The magnitude of the difference was also surprising — on the order of 0.5% — which is quite large when one considers that the average baseline HbA 1c was 7.5% in our study population and that being 'controlled' on treatment is defined as an HbA1c level less than 7.0%. Therefore, the difference in treatment response was well within the range of influencing whether patients achieved diabetes 'control,' " Dr. Williams told Medscape Medical News.

Overall response to metformin increased with increasing baseline HbA1c levels (P < .001), and that effect was stronger among African Americans than among white patients (P < .001).

In another analysis that accounted for the full duration between HbA1c measurements — 9.1 months for African Americans vs 8.3 months for whites — and after adjustment for confounders, the drop in HbA1c from baseline was lower (0.22 percentage points overall), but still significantly greater for the African American group (0.26 vs 0.17 percentage points).

The greater response to metformin among African Americans remained in other sensitivity analyses, including those restricting the analysis to patients who had de novo treatment of diabetes with metformin monotherapy, a baseline HbA1c measurement within 2 weeks preceding treatment initiation, and a follow-up HbA1c measurement between 4 and 5 months after metformin initiation.

Areas of Future Study

Dr. Williams told Medscape Medical News that because he and his colleagues carefully accounted for baseline differences such as body weight, age, sex, and treatment adherence, the differences were unlikely to be related to those factors.

"Therefore, genetic influences may very well underlie the difference — or part of the difference — in metformin treatment response, but this will require more research. As we learn more about the genetics of drug response, we are finding a number of circumstances in which genetic differences can influence both the likelihood of responding and the likelihood of having a severe side effect to medication, [such as with] warfarin and carbamazepine."

He added, "This work highlights the importance of assessing medication response in all populations, as groups may have different rates of treatment-related benefits and harms. These between-group differences can then be leveraged to identify both the genetic and nongenetic predictors of medication response."

This study was funded by grants from the National Institutes of Health and the Fund for Henry Ford Hospital. Dr. Williams and his coauthors have disclosed no relevant financial relationships .

J Clin Endocrinol Metab. Published online June 12, 2014. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: