Race, ethnicity, and use of thiazolidinediones among US adults with diabetes

Objective: Significant race and ethnic disparities exist in diabetes-related health care. Using a nationally representative database, we sought to determine if use of thiazolidinediones (TZDs) differs by race and ethnicity. As a secondary objective, we sought to determine if race and ethnicity is associated with use of older oral antidiabetic agents, such as sulfonylureas and metformin. Research design and methods: Adult respondents to the 2003 Medical Expenditure Panel Survey with diabetes, identified by diagnosis code or self-report, were included. Race/ethnic groups were defined as: White/not-Hispanic; Black/not-Hispanic; Hispanic; or Other/not-Hispanic. Associations between use of oral antidiabetic agents (defined as 2 1 prescription for a TZD, sulfonylurea, or metformin) and race/ethnicity, sex, age, insurance status, poverty status, and having a usual source of care were evaluated in univariate analyses with chi(2) tests and in adjusted analyses using logistic regression methods for survey data. Results: A total of 1873 US adults with diabetes were identified, with use of oral antidiabetic agents varying by drug class: 23.1% received TZDs, 45.3% received metformin, and 43.8% received sulfonylureas. Use of oral antidiabetic agents, by drug class, did not differ significantly by race/ethnicity (p = 0.33 for TZDs, p = 0.43 for metformin, p = 0.38 for sulfonylureas). In univariate analyses, only insurance status was significantly associated with use of TZDs (p = 0.03), and no variables were associated with use of sulfonylureas or metformin. In adjusted logistic regression analyses, there were no significant predictors of the use of TZDs or metformin, and only age was significantly associated with the use of sulfonylureas. Conclusions: In a nationally representative database, fewer US adults with diabetes received TZDs compared with sulfonylureas or metformin in 2003. Although we were not able to differentiate between type 1 and type 2 diabetes, nor did we assess oral agent monotherapy versus combination therapy, we found that use of TZDs, sulfonylureas, and metformin did not differ based on race/ethnicity or other demographic variables such as sex, insurance status, poverty status, or having a usual source of health care.