ANN ARBOR, Mich. -- Despite decades of advances in diabetes care, African Americans and Latinos are still far less likely than whites to have their blood sugar under control, even with the help of medications, a new nationally representative study finds. That puts them at a much higher risk of blindness, heart attack, kidney failure, foot amputation and other long-term diabetes complications.
The comprehensive new national study of middle-aged and older adults, published in the Sept. 24 issue of the Archives of Internal Medicine, was performed by a team from the University of Michigan and the VA Ann Arbor Healthcare System.
The study documents the persistence of strong racial and ethnic disparities in diabetes control, which have been observed for decades and contribute to the much greater impact of diabetes on those two ethnic groups. The results suggest that diabetes will continue to kill and disable black and Latino adults disproportionately for decades to come.
But the study delves deeper into the reasons behind this difference in blood sugar levels, using complex statistical analysis to find factors that do -- and don't -- play a role. For instance, diabetes control was worse among black and Latinos under age 65.
Most notably, two factors were found to account for a sizable portion of the racial and ethnic difference in glucose control: how well patients persist in taking their diabetes medicines regularly, and how they respond emotionally to having diabetes. Fortunately, these factors are likely to change in response to specific outreach efforts -- including some now underway by the U-M researchers. The study also hints that more factors are at work.
"While we were taken aback to see that diabetes control still varies so much by race and ethnicity, we're encouraged that two of the crucial factors are modifiable," says Michele Heisler, M.D., MPA, an assistant professor of Internal Medicine at the U-M Medical School and a research scientist at the VA Ann Arbor's Center for Clinical Practice Management Research. "To improve diabetes outcomes, we must do better at supporting all patients in managing their disease through treatment and lifestyle change. But we need to tailor specific interventions to address the barriers to achieving good diabetes control that African American and Latino adults with diabetes disproportionately face."
The study is based on very recent data from the Health and Retirement Study, a decades-long national effort to assess the health of adults over age 50 through regular completion of intensive questionnaires and health examinations.
Funded by the National Institute on Aging, and based at the U-M Institute for Social Research, the HRS began assessing the blood sugar levels of participants in 2003. In the older age groups where Type II diabetes is mostly found, the new study is larger than the other major source of population-wide data on this issue, the National Health and Nutrition Examination Survey (NHANES) run by the Centers for Disease Control and Prevention.
In all, 1,199 people over age 55 with diabetes were included in the new study. Their blood sugar was measured using the A1C test, which gives an average blood glucose level over the last three months and is considered a more accurate gauge of glycemic control than a simple glucose test.
"The ability to obtain such an important clinical marker on a large national sample is a major step forward in using population surveys to understand health disparities in the older population," said David Weir, Ph.D., director of the Health and Retirement Study and a research professor at ISR.
Current guidelines call for people with diabetes to maintain an A1C level of under 7 percentage points, to slow the rate of damage to nerves, blood vessels and organs that can lead to deadly and debilitating diabetes complications. People without diabetes typically have an A1C under 6 points.
But when the researchers analyzed data from study participants who were taking medications to control their blood sugar, the difference between the mean A1C for whites and the means for the other ethnic groups was large. White people had a mean A1C of 7.22 points, while the levels for blacks and Latinos were 8.07 and 8.14, respectively. People with diabetes are typically prescribed medications for glucose control only when diet and exercise no longer keep their levels in check.
An even bigger difference was seen when the researchers looked at the 286 participants on medications who were between ages 55 and 64 - too young for Medicare coverage. Whites had an average A1C of 7.46, but blacks were at 8.96 and Latinos were at 8.91. By contrast, there was a much smaller difference in average A1C among members of the three groups over age 65.
The researchers then performed a statistical analysis that took into account all of the available information about all the participants who were taking medication -- everything from their education level and annual household income to their mental health, insurance coverage status, quality of health care, medication regimens, exercise, diet, as well as their attitudes and behaviors about taking medications, monitoring blood sugar levels, and other key diabetes self-care tasks. The data also included participants' answers to a questionnaire that assesses a person's emotional response to living with diabetes, and a questionnaire about how they were managing their disease - including how well they adhered to the diabetes medications prescribed by their doctors.
A multvariate statistical analysis then allowed the researchers to separate out factors associated with higher A1C levels, and to assess how those factors in turn were associated with ethnicity. It also allowed them to adjust for differences in income, education, and all the other factors.
In the end, the factors that showed the strongest influence on racial and ethnic differences in A1C levels were medication adherence (especially among African Americans) and emotional distress related to diabetes (especially among Latinos). African Americans reported more barriers to taking their medications, and less adherence to their medication, than the other groups. Meanwhile, Latinos reported much higher levels of distress related to their diabetes than other groups.
Even so, all the factors examined in the analyses that might account for the observed racial and ethnic disparities in glycemic control accounted for only 14 percent of the African American-white disparity, and 19 percent of the Latino-white disparity, in blood sugar control. Meanwhile, differences in income and education level - two factors long hypothesized to be key determinants of worse diabetes outcomes - did not explain the glucose control differences, once the other factors were included in the analyses.
The authors conclude that additional factors not assessed in the study, such as genetics, stress levels and other environmental factors, intensity of medication regimens, and the generosity of patients' prescription drug insurance coverage must account for a large part of the picture.
"Medication adherence was one of the strongest predictors of glucose control across the board," says Heisler. "This reinforces that by targeting barriers to medication adherence -- such as patient-doctor communication about medications, patient trust in health systems, patient confidence that medication actually helps, cost barriers, and other barriers that African Americans disproportionately face -- we can make a difference."
"Diabetes is one of the most important health challenges faced by Americans and American society today," notes Richard Suzman, Ph.D., director of behavioral and social research at the National Institute on Aging. "These results illuminate some of the behavioral and other issues associated with glycemic control that can be useful in designing strategies and interventions to reach diverse populations."
Heisler and her colleagues are currently conducting two randomized controlled trials of such interventions in people with diabetes who have high A1C levels, blood pressures, and lipid (cholesterol) levels. One, supported by the National Institutes of Health and the VA, includes nurse-led group sessions where patients can break their longer-term diabetes self-care goals into short-term specific steps, and chance for patients to link up with a diabetes peer "buddy" who faces similar self-care challenges, to provide mutual coaching and support during weekly telephone calls.
The other, funded by the National Institute for Diabetes and Digestive and Kidney Diseases and VA, is training VA pharmacists to reach out to diabetes patients with poor risk factor control and pharmacy data that shows difficulties refilling medications. The clinical pharmacists will provide "motivational-interviewing-based" adherence assessment and counseling. This proactive outreach will specifically target blood pressure, which like glucose is a crucial factor in the development and progression of diabetes complications. The pharmacists will also have the ability to increase patients' dosages of blood pressure medications, within a framework pre-approved by physicians.
In addition to Heisler and Weir, the newly published study is co-authored by U-M and VA researchers Jessica Faul, MPH, Rodney Hayward, M.D., Kenneth Langa, M.D., Ph.D., and Caroline Blaum, M.D. It was supported by NIA, VA, and the Michigan Diabetes Research and Training Center.
Archives of Internal Medicine, Vol. 167 No. 17, Sept. 24, 2007.
Note to broadcasters:
Video and audio are available upon request