No. 3, July 2006
Racial Disparities in Diabetes-related Preventive Care: Results From the Missouri Behavioral Risk Factor Surveillance System
Joseph William LeMaster, MD, MPH, Fungai Chanetsa, PhD, Julie M. Kapp, PhD, MPH, Brian M. Waterman, MPH
Suggested citation for this article: LeMaster JW, Chanetsa F, Kapp JM, Waterman BM. Racial disparities in diabetes-related preventive care: results from the Missouri Behavioral Risk Factor Surveillance System. Prev Chronic Dis [serial online] 2006 Jul [date cited]. Available from: URL: http://www.cdc.gov/pcd/issues/2006/
Racial disparities exist in the rates of diabetes complications in the United States and in the state of Missouri. It is unclear to what degree such disparities involve diabetes-related preventive care. We sought evidence for racial disparities in diabetes-related preventive care between non-Hispanic blacks and whites in Missouri.
We analyzed data from the Missouri Behavioral Risk Factor Surveillance System from 1994 through 2002. This state-specific survey is conducted annually among a representative sample of Missourians. We examined data from 842 Missourians who reported a diagnosis of
type 1 or type 2 diabetes and who had consulted a health professional in the 12 months before they were interviewed. We analyzed
reported receipt of glycosylated hemoglobin testing, foot examinations, and dilated eye examinations in the year
Non-Hispanic blacks were significantly less likely than whites to report having
had glycosylated hemoglobin testing (odds ratio [OR], 0.47; 95% confidence interval [CI], 0.22–0.99) but more likely to
report having received foot examinations (OR, 1.99; 95% CI, 1.21–2.39). There was no difference between blacks and whites in the probability of dilated eye examinations (OR, 1.49; 95% CI, 0.94–2.36).
Compared with whites, non-Hispanic blacks in Missouri receive adequate screening for diabetic complications but not for glycemic control. Further studies are needed to investigate whether these disparities are linked to differences in the rate of diabetes complications in Missouri.
Back to top
In the last decade, concerted efforts have been made to increase regular
screening for diabetes complications to improve health outcomes. The American
Diabetes Association recommends annual foot and dilated eye examinations; measurement of
glycosylated hemoglobin (HbA1c), lipids, creatinine,
cholesterol, and urine protein; and biannual measurement of blood pressure (1). These recommendations have led to considerable increases in the use of these tests
in the last decade (2-4) and associated decreases in the rate of diabetes complications (2-6). Because provision of these diabetes preventive examinations is monitored by many large clinics (3,7,8), by the U.S. Department of Veterans
Affairs (9), and by population-based surveys, including the Medicare Current Beneficiary Survey (10), the National Health and Nutrition Examination Survey (NHANES) (2), and the Behavioral Risk Factor Surveillance System (BRFSS) (11), these measures make it possible to compare diabetes quality of care across diverse populations and health care systems.
U.S. incidence rates for diabetes mellitus and complications from diabetes
are higher for racial minority groups than for whites (10,12-20). Racial
disparities in routine preventive diabetes care may contribute to these higher
rates (10,12,13,21-24). Non-Hispanic black adults with diabetes who receive
health care from U.S. managed care organizations are reportedly less likely than
whites to receive annual HbA1c or eye screening (13,21). Reports from nationally
representative U.S. population data are less consistent: one study analyzing
Medicare claims data from more than 300,000 beneficiaries nationwide found that black Medicare beneficiaries are about a third less likely than whites to have HbA1c measured or to be referred for eye examinations (22). In contrast, reports analyzing data
from Veterans Affairs facilities, the NHANES, and the BRFSS have not found racial differences in HbA1c testing, eye care, insulin use, or blood pressure treatment for individuals with diabetes (5,23,25). These findings suggest that some U.S. populations may experience racial disparities in diabetes care, whereas others do not.
In some U.S. states, large racial disparities exist in the rates of diabetes complications.
From 1992 to 2002 in Missouri, diabetes-related deaths among non-Hispanic blacks occurred at twice the rate among whites, and emergency department visits for diabetes-related complications among blacks were about four times more frequent than among whites (26). Given that diabetes
preventive care reduces diabetes-related complications and mortality, we hypothesized that diabetes preventive care would be less frequently received by non-Hispanic blacks than by whites in Missouri. We used state-specific data collected from the Missouri BRFSS from 1994 through 2002 to determine whether preventive diabetes care in Missouri differed by race and whether these differences changed
Back to top
Data source and study population
The U.S. Centers for Disease Control and Prevention (CDC) collects annual, state-specific data for the BRFSS in all 50 states and the District of Columbia. The Missouri BRFSS uses a disproportionate stratified sampling methodology to select respondents from seven state health regions including
the city of St Louis. Data are collected through telephone interviews among a sample of noninstitutionalized adults (aged 18 years and older) living in households with a telephone (27).
In this analysis, we included 842 Missourians who reported having a medical diagnosis of diabetes mellitus and non-Hispanic black or white race, who consulted a physician or health professional for a routine checkup in the year
before being interviewed, and who answered questions about their diabetes preventive care during that period.
Interviews were conducted in 1994, 1996, 1997, 2000, and 2002. We compiled data
for these survey years to determine whether there was any trend in the frequency of
reported examinations. We excluded other racial or ethnic groups that comprised only 4% of the total BRFSS sample
because the small sample size precluded statistical analysis.
The BRFSS includes questions about preventive health practices and risk behaviors that are linked to chronic diseases, injuries, and preventable infectious diseases in the adult, noninstitutionalized civilian population. The set of questions varies somewhat among states in a given year and from year to year in a given state. Our analysis focused on a common set of questions about
diabetes-related preventive care services (outcomes) that were reported in Missouri during the
years we analyzed: whether respondents reported that, in the 12 months before being interviewed, a health professional had conducted a
“hemoglobin A1c” blood test, checked their feet for any sores or irritations, or performed an eye examination in which the pupils were dilated. To focus our analysis on the
quality of health care received by respondents with diabetes, we included only respondents who had
been to a health professional in the 12 months before being interviewed. These data were gathered for the Missouri BRFSS in
each of the survey years included in this study. Missouri BRFSS survey methods and
wording of questions about the diabetes preventive care outcomes we analyzed did not change
substantially during this period. For the years analyzed, the Missouri BRFSS included questions about cholesterol and blood pressure screening in 1997 only, so these questions were not included in our analysis.
The BRFSS data are weighted by CDC to correct for the differences in probability of selection. The weighting takes into account the number of adults in the household, the number of telephones per household, and telephone coverage. In addition, poststratification weighting accounts for nonresponse by age, race, and sex and is based on the most recent census data in each U.S. state. This
weighting allows inferences about the state population to be made from the sample (27). In all of our analyses, we adjusted and applied these weights to account for the sampling scheme in Missouri.
We conducted descriptive analyses initially using two-way contingency tables. We used multivariate logistic regression to estimate the association between race (non-Hispanic blacks
compared with whites) and preventive care outcomes. We tested whether other respondent characteristics individually or jointly confounded these associations using likelihood ratio tests in these models. Characteristics were
included in the final model for each outcome as adjustment variables if they changed the crude odds ratios (ORs) describing the effect of race on the preventive care outcome under consideration by more than 10%. We adjusted all models for income, employment, and enrollment in either a public or private health insurance plan. We also used likelihood ratio tests to assess interactions
between race and other respondent characteristics (including the year respondents were interviewed). Item nonresponse was minimal except for income, which was not reported by 14% of respondents. We used linear regression to impute income when these data were missing. We replaced missing income values with those estimated by a multivariate linear regression model that estimated income as a
function of all other variables that we included in the preventive care models. We then refit the final preventive care models using this imputed income variable. To determine the degree to which nonresponse by
non-Hispanic black respondents may have led to selection bias in our analysis, we used a logistic regression model to estimate the association between race and nonresponse for each preventive care
outcome that was asked by the BRFSS in the years we examined. We adjusted those estimates for sex, age, income, education, employment status, health insurance coverage, marital status, body mass index (BMI), and the year respondents were interviewed. We assessed model calibration for all models with the Hosmer-Lemeshow test (28). All data management and analyses were done with SAS,
version 8.02 (SAS Institute Inc, Cary, NC). We used an α level of .05 to define significance. The study was approved by the University of Missouri Institutional Review Board.
Back to top
Non-Hispanic blacks represented 16% of the Missouri population with diabetes who were sampled during the years we examined
(Table 1). These respondents were younger,
had slightly less income, and were less likely to be married than white respondents. Sex, education, BMI, health coverage, and employment did not differ by race.
Among the sample, 66% reported having had their feet checked by a health professional in the past year, with proportionally higher rates among
(Table 2). Crude ORs indicated that blacks were more likely than whites to have had their feet checked in the past year. After adjustment for employment,
enrollment in health insurance, BMI, and income, this difference remained significant (OR,
1.99; 95% confidence interval [CI], 1.21–2.39).
Sixty-seven percent of respondents reported having had a dilated eye examination in the past year, with a higher proportion reporting this outcome among blacks (72%) than whites (68%) (Table 2).
Although the crude OR indicated that blacks were more likely to have had an eye examination than whites, this difference was not statistically significant. After the model was adjusted for health insurance
coverage, income, age, and employment, the increased OR for blacks receiving an eye examination remained nonsignificant (OR, 1.49; 95% CI, 0.94–2.36).
Among the sample, 90% reported having had at least one HbA1c test in the past year. Blacks were less likely to report having received this test than whites in the crude analysis (Table 2). After adjustment for health insurance
coverage, employment, and income, this difference remained significant (OR, 0.47; 95% CI, 0.22–0.99).
There was no evidence that blacks were selectively excluded from the models because of nonresponse
(Table 3). After the model was adjusted for respondent characteristics and the year of interview,
non-Hispanic blacks were more likely to respond to questions about foot examination, whereas there was no difference in nonresponse between whites and blacks for the eye examination and HbA1c models.
There was no evidence of interaction between race and any of the adjustment variables in the regression models. There was also no evidence for increasing or decreasing trend by year for any of the diabetes preventive care outcomes that we considered (data not shown). All statistical models described revealed appropriate (i.e., nonsignificant) calibration based on the Hosmer-Lemeshow
Back to top
This study suggests that there was substantial underuse of a key test of
diabetes control, HbA1c level, among non-Hispanic black Missourians with
diabetes from 1994 to 2002. Blacks reported adequate preventive care for at least
two potential diabetes complications, retinopathy and neuropathy, at least as
frequently as whites; foot examination was reported more frequently by blacks, and
eye care did not differ by race. Our finding of a disparity in HbA1c testing
differs from a previous national BRFSS analysis, which found that non-Hispanic blacks
did not differ from other groups in receipt of preventive services (5).
Our results cannot be explained by differences in access to medical care between blacks and whites,
because all respondents had been examined by a physician in the last year. We included in our analysis only respondents who had consulted a health care professional, because the BRFSS does not include questions about why respondents thought they did or did not have access to health care
professionals. Although people with diabetes may not receive diabetes preventive care — either because they do not have access to a care provider or because the provider does not deliver the care — an analysis exploring reasons why
individual respondents did not consult a health care professional was beyond the scope of this study.
Low-income patients and patients without health insurance are less likely to be up-to-date on preventive services of all types (29); however, all of our regression models adjusted for employment, income, and enrollment in health insurance. Other factors related to the health care encounter may also have an influence; U.S. physicians are under pressure to see more patients in less time
(24), a situation that reduces time spent on diabetes preventive care (30). In one survey study of physicians,
they reported that their biggest challenge to adequate treatment of diabetes was not patients’ race;
it was inadequate time and resources (31).
Differences in patients’ and health care providers’ perceptions about the importance of diabetes in the patient’s life may strongly influence providers’ delivery of preventive diabetes care and patients’ achievement of optimal glycemic control (31,32). Physicians in one large U.S. health maintenance organization stated that there seemed little reason to order
HbA1c tests for black patients because these tests had been consistently high in the recent past (i.e., reflecting poor glycemic control) (21). Similar practices among Missouri physicians may have contributed to our finding that blacks reported
having had fewer HbA1c tests than whites in that state. In contrast, blacks may have reported that they
had received foot and eye examinations at least as frequently as whites
because both providers and patients perceived correctly that diabetes foot and eye complications occur frequently among blacks (and thus are important to detect early). The alternative explanation, that black respondents preferentially remembered foot examinations better and HbA1c tests worse than white respondents, seems less reasonable.
Black patients depend heavily on the medical system to help them manage their diabetes as they grow older because they realize that their risk of a grave complication is high (33). Non-Hispanic black, Hispanic, and white patients all attend ophthalmologic, podiatric, and weight reduction clinics equally when referred
by their primary care physician (21). Although black
diabetes patients make fewer outpatient physician visits and use emergency services more than whites (10), we are unaware of any published evidence that suggests that
they are indifferent about receiving diabetes preventive care.
Given the relatively high baseline prevalence of the outcomes examined in this study, the adjusted ORs we report,
although valid and sound measures of association, may not approximate relative risk to the extent that they would if the baseline prevalence of the outcomes was low. Although other methods of presenting findings are available (i.e., model-adjusted probabilities), we chose to report model findings in the more common format of model-adjusted ORs.
This study has several limitations. First, results from the BRFSS may not equally represent all black residents in Missouri. The BRFSS does not collect data on people without telephones. Although our adjusted BRFSS weights accounted for this at the state level, blacks from rural areas of Missouri, where telephones
are less common than in urban areas, may have been underrepresented
(27). On the other hand, our analysis may have overestimated preventive service use because individuals who
cannot afford a telephone are also less likely to receive regular diabetes preventive care (34). Second, the BRFSS elicits responses about medical care from patients rather than from health care providers. Self-reported data are subject to recall bias. Differences in socioeconomic
status or cultural or language barriers may have differentially affected reporting patterns for blacks compared with whites. Third, because the data from respondents from other ethnic groups were sparse, our analysis only compared non-Hispanic blacks to whites; however, similar observations have been made about the difference in quality of care between Hispanics and non-Hispanics. In fact,
Hispanics have less frequent HbA1c testing than non-Hispanic blacks (21). Fourth,
numerous responses to the HbA1c question were missing, making the
confidence intervals for that outcome quite wide. Fifth, nonresponse to diabetes preventive care questions was high for the HbA1c
test model; however, we found no association between race and nonresponse for that outcome. Sixth, to examine trends in diabetes care over time, we limited the number of outcomes that we analyzed to those that were repeatedly included in the Missouri BRFSS during the years we studied. Although we cannot generalize these results to other preventive care practices, these findings are consistent with what others have found (10,12,13,17,21,35,36). Finally, because the BRFSS only
surveys noninstitutionalized people, our results cannot be generalized to people who live in nursing homes, prisons, or other institutions.
We cannot conclude, based on our results, that disparities in HbA1c testing lead to differences in rates of diabetes-related complications and mortality in Missouri, because the BRFSS
does not collect information on the occurrence of diabetes complications;
however, the results are suggestive. In Michigan, non-Hispanic blacks with diabetes in a health maintenance organization had higher
HbA1c values and poorer renal function than whites, and blacks did not receive recommended levels of preventive diabetes care (37). In another Missouri survey, black youths with diabetes had substantially poorer glycemic control than their white counterparts, although that study did not investigate the association between glycemic control and the quality of medical care that
respondents received (38). Another recent survey of non-Hispanic blacks in Missouri revealed that respondents did not feel that they received the same quality of health care as whites (26). Further studies should investigate the association between racial disparities in diabetes preventive care and the occurrence of complications from diabetes in
midwestern U.S. states.
Back to top
This project was funded by the Center for Health Care Quality, Department of Health Management and Informatics, at the University of Missouri–Columbia in Columbia, Mo. The authors acknowledge the support of the Missouri BRFSS data collection team and Dr David Mehr, who gave helpful editorial suggestions
for initial drafts of the manuscript.
Back to top
Corresponding Author: Joseph William LeMaster, Department of Family and Community Medicine, University of Missouri–Columbia School of Medicine, M241A Health Sciences
Bldg DC032.00, Columbia, MO 65212. Telephone: 573-882-1584. E-mail: firstname.lastname@example.org.
Author Affiliations: Fungai Chanetsa, Northwest Missouri State University, Maryville, Mo; Julie M. Kapp, Department of Health Management and Informatics, University of Missouri–Columbia, Columbia, Mo; Brian M. Waterman, Waterman Research Solutions, St Louis, Mo.
Back to top
- American Diabetes Association.
Standards of medical care in diabetes. Diabetes Care 2004;27(Suppl 1):S15-35.
- Harris MI.
Health care and health status and outcomes for patients with type 2 diabetes. Diabetes Care 2000;23(6):754-8.
- Huang ES, Gleason S, Gaudette R, Cagliero E, Murphy-Sheehy P, Nathan DM, et al.
Health care resource utilization associated with a diabetes center and a general medicine clinic. J Gen Intern Med 2004;19(1):28-35.
- Meigs JB, Stafford RS.
Cardiovascular disease prevention practices by U.S.
physicians for patients with diabetes. J Gen Intern Med 2000;15(4):220-8.
- Saaddine JB, Engelgau MM, Beckles GL, Gregg EW, Thompson TJ, Narayan KM.
A diabetes report card for the United States: quality of care in the 1990s. Ann Intern Med 2002;136(8):565-74.
The effect of intensive diabetes therapy on the development and progression of neuropathy. The Diabetes Control and Complications Trial Research Group. Ann Intern Med 1995;122(8):561-8.
- Parchman ML, Burge SK; Residency Research Network of South Texas Investigators.
Continuity and quality of care in type 2 diabetes: a Residency Research Network of South Texas study. J Fam Pract 2002;51(7):619-24.
- Wagner EH, Sandhu N, Newton KM, McCulloch DK, Ramsey SD, Grothaus LC.
Effect of improved glycemic control on health care costs and utilization. JAMA 2001;285(2):182-9.
- Boyko EJ, Ahroni JH, Stensel V, Forsberg RC, Davignon DR, Smith DG.
A prospective study of risk factors for diabetic foot ulcer. The Seattle Diabetic Foot Study. Diabetes Care 1999;22(7):1036-42.
- Chin MH, Zhang JX, Merrell K.
Diabetes in the African-American Medicare population. Morbidity, quality of care, and resource utilization. Diabetes Care 1998;21(7):1090-5.
- Centers for Disease Control and Prevention (CDC).
Levels of diabetes-related preventive-care practices — United States, 1997-1999. MMWR Morb Mortal Wkly Rep 2000;49(42):954-8.
- Gary TL, Narayan KM, Gregg EW, Beckles GL, Saaddine JB.
Racial/ethnic differences in the healthcare experience (coverage, utilization, and satisfaction) of US adults with diabetes. Ethn Dis 2003;13(1):47-54.
- Gary TL, McGuire M, McCauley J, Brancati FL.
Racial comparisons of health care and glycemic control for African American and white diabetic adults in an urban managed care organization. Dis Manag 2004;7(1):25-34.
- Brancati FL, Whittle JC, Whelton PK, Seidler AJ, Klag MJ.
The excess incidence of diabetic end-stage renal disease among blacks. A population-based study of potential explanatory factors. JAMA 1992;268(21):3079-84.
- Roseman JM. Diabetes in black Americans. In: Harris MI, Hamman RF, editors. Diabetes in America. Washington (DC): U.S. Government Printing Office; 1985. p. 1-24.
- Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE, Little RR, et al.
Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988-1994. Diabetes Care 1998;21(4):518-24.
- Cowie CC, Port FK, Wolfe RA, Savage PJ, Moll PP, Hawthorne VM.
Disparities in incidence of diabetic end-stage renal disease according to race and type of diabetes. N Engl J Med 1989;321(16):1074-9.
- Miller AD, Van Buskirk A, Verhoek-Oftedahl W, Miller ER.
Diabetes-related lower extremity amputations in New Jersey, 1979 to 1981. J Med Soc N J 1985;82(9):723-6.
- Resnick HE, Valsania P, Phillips CL.
Diabetes mellitus and nontraumatic lower extremity amputation in black and white Americans: the National Health and Nutrition Examination Survey Epidemiologic Follow-up Study, 1971-1992. Arch Intern Med 1999;159(20):2470-5.
- Lavery LA, van Houtum WH, Ashry HR, Armstrong DG, Pugh JA.
Diabetes-related lower-extremity amputations disproportionately affect Blacks and Mexican Americans. South Med J 1999;92(6):593-9.
- Martin TL, Selby JV, Zhang D.
Physician and patient prevention practices in NIDDM in a large urban managed-care organization. Diabetes Care 1995;18(8):1124-32.
- Schneider EC, Zaslavsky AM, Epstein AM.
Racial disparities in the quality of care for enrollees in medicare managed care. JAMA 2002;287(10):1288-94.
- Heisler M, Smith DM, Hayward RA, Krein SL, Kerr EA.
Racial disparities in diabetes care processes, outcomes, and treatment intensity. Med Care 2003;41(11):1221-32.
- Lindblad CI, Hanlon JT, Artz MB, Fillenbaum GG, McCarthy TC.
Antidiabetic drug therapy of African-American and white community-dwelling elderly over a 10-year period. J Am Geriatr Soc 2003;51(12):1748-53.
- Harris MI.
Racial and ethnic differences in health care access and health outcomes for adults with type 2 diabetes. Diabetes Care 2001;24(3):454-9.
- Minority health disparities in Missouri. Data Book 1: African Americans. St Louis: Missouri Foundation for Health;[updated 2004 Dec 31; cited 2005 Jun 1]. Available from: URL:
- Centers for Disease Control and Prevention. Behavioral Risk Factor Surveillance System
operational and user's guide. Version 3.0. Atlanta (GA): Centers for Disease Control and Prevention;[updated 2005 Mar 4; cited 2005 Mar 29]. Available from: URL:
- Hosmer DW Jr, Lemeshow S. Assessing the fit of the model. In: Applied logistic
2nd ed. New York (NY): John Wiley and Sons; 2000. p. 147-155.
- Solberg LI, Brekke ML, Kottke TE.
Are physicians less likely to recommend preventive services to low-SES patients? Prev Med 1997;26(3):350-7.
- Streja DA, Rabkin SW.
Factors associated with implementation of preventive care measures in patients with diabetes mellitus. Arch Intern Med 1999;159(3):294-302.
- Egede LE, Michel Y.
Perceived difficulty of diabetes treatment in primary care: does it differ by patient ethnicity? Diabetes Educ 2001;27(5):678-84.
- Freeman J, Loewe R.
Barriers to communication about diabetes mellitus. Patients' and physicians' different view of the disease. J Fam Pract 2000;49(6):507-12.
- Reid BV.
"It's like you're down on a bed of affliction": aging and diabetes among black Americans. Soc Sci Med 1992;34(12):1317-23.
- Centers for Disease Control and Prevention (CDC).
Preventive-care practices among persons with diabetes — United States, 1995 and 2001. MMWR Morb Mortal Wkly Rep 2002;51(43):965-9.
- Carrasquillo O, Himmelstein DU, Woolhandler S, Bor DH.
Going bare: trends in health insurance coverage, 1989 through 1996.
[Published erratum in: Am J Public Health 1999;89(2):256]. Am J Public Health 1999;89(1):36-42.
- Weinick RM, Zuvekas SH, Cohen JW.
Racial and ethnic differences in access to and use of health care services, 1977 to 1996. Med Care Res Rev 2000;57 Suppl 1:36-54.
- Wisdom K, Fryzek JP, Havstad SL, Anderson RM, Dreiling MC, Tilley BC.
Comparison of laboratory test frequency and test results between African-Americans and Caucasians with diabetes: opportunity for improvement. Findings from a large urban health maintenance organization. Diabetes Care 1997;20(6):971-7.
- Auslander WF, Thompson S, Dreitzer D, White NH, Santiago JV.
Disparity in glycemic control and adherence between African-American and Caucasian youths with diabetes. Family and community contexts. Diabetes Care 1997;20(10):1569-75.
Back to top