OBJECTIVE—There is conflicting evidence about whether nonwhite Americans with diabetes have an increased risk of cardiovascular disease (CVD). Because geographic region is known to influence the risk of CVD in the U.S., we sought to determine the effects of race and region on cardiovascular morbidity among elderly Americans with diabetes.

RESEARCH DESIGN AND METHODS—We performed a national, retrospective, cohort study using the Medicare claims of 126,153 white and 17,962 black patients with diabetes, aged ≥65 years in 1994, who were followed through 1999 for incident acute myocardial infarction, ischemic heart disease, stroke, and heart failure. The effect of race, sex, and region on the incidence of these diseases was assessed using Cox proportional hazards regression, adjusting for baseline demographics and comorbidities.

RESULTS—The incidence of any CVD ranged from 23.9/100 person-years among southern black men to 29.2/100 person-years among non-southern black women. The risk of CVD was lower among southern black men (hazard ratio 0.87 [95% CI 0.82–0.92]) and women (0.95 [0.91–0.99]) than their southern white counterparts. In the three other U.S. regions combined (northeast, midwest, and west), black men had a similar risk for CVD (1.01 [0.95–1.07]), and black women had a greater risk (1.10 [1.05–1.16]) than non-southern white men and women, respectively.

CONCLUSIONS—Among elderly Americans with diabetes, the incidence of CVD is unlikely to differ a great deal between whites and blacks. Residence in the South seems to confer a modest benefit for elderly black people with diabetes.

Race and ethnicity significantly influence the risk of cardiovascular disease (CVD) in the U.S. For example, black race is associated with increased mortality from ischemic heart disease (IHD) and stroke (13) compared with whites. Diabetes, a major risk factor for CVD (4), affects black Americans disproportionately, and frequently, diabetes is not as well controlled in blacks as in whites (58). One might suspect, therefore, that blacks with diabetes would have excess CVD. Although one U.S. study (9) reported greater CVD prevalence among black geriatric patients with diabetes, others (1012) have not found excess CVD. In fact, blacks of African-Caribbean descent in the U.K. who have diabetes have been found to be at a lower risk of CVD than white Europeans with diabetes (13). We previously found that heart failure incidence was similar between elderly blacks and whites with diabetes (14), but neither that study nor the other U.S. studies cited have considered whether among individuals with diabetes the effect of race on CVD outcomes differs with geography. Cardiovascular mortality varies across the U.S., with the South having the highest rates for stroke and CVD mortality regardless of the racial group studied (15,16). The burden of diabetes also varies by region, with the South having the highest prevalence (17). Accordingly, we hypothesized that the effect of race on risk of CVD among individuals with diabetes may vary by region. This report investigates, among the elderly with diabetes, black-white differences in the incidence of acute myocardial infarction (AMI), IHD, cerebrovascular accident (CVA; including both stroke and transient ischemic attack), and heart failure and evaluates the extent to which region influences any racial differences observed.

Medicare is the government-financed medical insurance program for the elderly and disabled who contributed to the Social Security program in the U.S.; virtually all Americans aged ≥65 years are covered. We obtained data from the 5% Standard Analytical File, which contained Medicare claims for 1,941,453 beneficiaries in 1994. Details regarding the data collection have been published (18). Race (by self-report at enrollment into Social Security), age, state of residence, and mortality information were as reported in the Medicare Denominator File. After excluding subjects aged <65 years and those older who died in 1994, there were 1,551,586 beneficiaries age ≥65 years alive through 1 January 1995. We further excluded 31,151 (2.0%) who did not live in the U.S., 92,818 (6.0%) who were not enrolled in both Medicare Part A (inpatient) and Part B (outpatient), and 142,964 (9.2%) who were enrolled in a health maintenance organization (HMO) in 1994 (and thus utilization claims are not available). Among the 1,284,653 remaining, individuals were classified as having diabetes if, in 1994, there were at least two Part B claims or one hospitalization claim with a diabetes-related ICD-9 diagnosis code (250.xx). This classification yielded 151,523 individuals with diabetes. Of this group, we excluded 4,404 subjects who identified themselves as Hispanic, Asian, Native American, or other race (their numbers were insufficient to compare them with whites by sex or region) and 3,004 with missing race/ethnicity data, resulting in a cohort of 126,153 white and 17,962 black older adults with diabetes. To assess representativeness, we compared the demographic characteristics of this cohort with the 1990 U.S. Census population aged ≥65 years (19).

Subjects were considered to have prevalent heart failure, AMI, IHD, or CVA if one or more inpatient or outpatient claims were found in 1994 with an ICD-9 code (online appendix [available at http://care.diabetesjournals.org]) indicating that disease; prevalent CVD was defined as one claim for any of these diagnoses. For a given incidence analysis, we first excluded subjects with that disease in 1994. We then identified the first CVD claim occurring after 1 January 1995. The date of CVD was the midpoint of the quarter in which the claim was filed, unless the participant also died in that quarter, in which case the date of CVD was the minimum of the midpoint of the interval or the date of death. Participants were censored if they died (censored on date of death) or joined an HMO (assessed annually, censored date 31 December of year before joining the HMO) before experiencing an end point. Follow-up extended until 31 December 1999. Comorbidities were defined as the presence, in 1994, of at least one claim containing a diagnosis code for hypertension, nephropathy, retinopathy, or neuropathy (online appendix); additionally, we identified individuals also enrolled in Medicare’s end-stage renal disease program. The U.S. was divided into four regions (south, west, midwest, and northeast) based on state groupings in the Census. For ease of presentation, we dichotomized the regions into south and “other regions” because preliminary analyses indicated that the major differences in patterns of CVD incidence were between the South and the other regions.

Analysis

χ2 Tests were used to assess racial differences in comorbidities and prevalence of CVD at baseline, and a Wilcoxon’s rank-sum test was used to compare the age distributions. We used survival analysis to calculate event rates, assuming person-time accrued from 1 January 1995 to the date of event or the censored date. Incidence rates were estimated as 100 times the number of events divided by the person-years of exposure. Kaplan-Meier methods were used to obtain 1- and 5-year disease-free survival estimates for blacks and whites separately for strata defined by region and sex. Log-rank tests were used to assess the effect of race on survival. Cox’s proportional hazards regression was used to calculate the adjusted hazard ratio for blacks versus whites for each incident CVD event. Models included interaction terms for race by geographic region and race by sex. A stratified analysis (yearly age strata up to age 84, then 85–89, and ≥90 years) was used to ensure unbiased estimates for the race effect (20). Given the fine stratification used, the results obtained were almost identical to those obtained from an alternative model using age as the time scale and stratifying by age at baseline. Other covariates included sex, region, and baseline values for comorbidities. All significance tests were two tailed. Analyses were performed using SAS (SAS 8.2; SAS Institute, Cary, NC).

Baseline characteristics

The sex distribution was similar, but the racial distribution differed from the characteristics of the U.S. population aged ≥65 years in the 1990 census (60% women; 89% white, 8% black, and 3% other). The characteristics of black and white Medicare beneficiaries meeting our selection criteria are presented in Table 1. Blacks were slightly younger than whites (mean age 73.8 vs. 74.5 years, P < 0.001). Renal disease, hypertension, and retinopathy were more common among blacks. Blacks were slightly more likely to have had a prior stroke, while whites were more likely to have had an AMI and IHD.

Incident CVD

After excluding subjects with prevalent disease, the median follow-up time was 2.6 years for IHD and 2.1 years for any CVD. The incidence rate of AMI was lower in blacks (4.0/100 person-years) than whites (4.8/100 person-years); a similar pattern was noted for IHD (blacks 15.7/100 person-years, whites 17.5/100 person-years) (both log rank P < 0.001). The rates were more similar for the other conditions (CVA, blacks 13.0/100 person-years, whites 12.8/100 person-years; heart failure, blacks 13.1/100 person-years, whites 12.9/100 person-years). When we examined rates by race and region, for most diagnoses we noted a lower rate for blacks, particularly women, in the South, compared with the other regions. Formal tests for interactions (see research design and methods) were significant for all diagnoses. The cumulative 5-year Kaplan-Meier incidence estimates of IHD by race, region, and sex (Fig. 1) are illustrative of this interaction, demonstrating differential incidence by race according to sex and region. Because of these interactions, stratified incidence rates are presented in Table 2. The incidence rates of AMI, IHD, and heart failure were generally lower among the four race/sex groups in the South than in their counterparts in the rest of the U.S. The incidence rate of CVA was higher in the South for white men and women and was lower in the South for black women. Southern blacks had lower rates of AMI, IHD, and CVA than southern whites. Southern black and white men and women have similar rates of heart failure. Additionally, black men in the other regions tend to have lower rates of disease than white men in the same regions, whereas black women in the other regions tend to have higher rates than the corresponding white women. The incidence of any CVD was slightly lower in the South.

Multivariate models

The results of the Cox’s proportional hazards regressions for CVD outcomes, simultaneously adjusting for comorbidities, age, sex, and region, are shown in Table 3. The comorbidities considered were significantly associated with an increased risk of each outcome. Because of the interactions present, the sex and region effects are presented by race; the race effect is presented by subgroups of region and sex.

Among whites, being male conferred a 6–28% increased risk of the individual outcomes and a 10% (95% CI 8–13) increased risk of any CVD outcome. In contrast, among blacks, being male conferred an increased risk only for AMI. Blacks living in the South (versus blacks in other regions) had a 9–28% lower risk of the individual CVD outcomes. Among whites, residence in the South was protective only against AMI (6% lower risk) and IHD (5% lower risk); in contrast, southern whites had a 9% higher risk of stroke. The effect of race in the South was, for black men, a 31% lower risk of AMI and a 26% lower risk of IHD than their white counterparts, while southern black women had a 20% lower risk of AMI and a 16% lower risk of IHD than southern white women. For any CVD event, black women in the South were at 5% lower risk than their white peers; the risk for southern black men was 13% lower than for whites. In contrast, in other regions, the effect of black race among women was an increased risk of IHD (11%), heart failure (6%), CVA (12%), or any CVD (10%). Among non-southern men, a lower risk among blacks was only seen for AMI (17%).

In a large cohort of elderly Medicare beneficiaries with diabetes who received care on a fee-for-service basis, we observed a substantial incidence of CVD among all sex/age/regional groups examined. We found some evidence supporting our hypothesis that there are regional differences in the effect of race on the risk of CVD among individuals with diabetes. Outside the South, black women had a small but significantly higher incidence than white women of any CVD and of IHD, heart failure, and CVA specifically; but inside the South, black women had a lower incidence of each CVD event (and of any CVD) than white women. Patterns for men also differed by region, as southern black men were always at less risk than white men for CVD, but outside the South there was a difference by race only for AMI (where black men were at lower risk).

Considerable data suggest that blacks with diabetes have worse control of glucose and blood pressure than whites (6,7,21). Worse control glucose of glycemia and blood pressure is associated with an increased risk of CVD events and suggest that there should be a much higher rate of CVD in blacks than in whites (22,23). Despite this, we did not find a substantially increased risk of CVD among black Medicare fee-for-service patients when we compared them with whites; as noted, in the South there was actually an advantage conferred by black race in all comparisons. We did find a modestly increased risk (range 6–12%) for black women in non-southern regions. This is in contrast to earlier reports utilizing data from one California HMO, the Veterans Health Administration, as well as national health statistics that found that blacks with diabetes were not at greater risk than whites (11,12,24,25). We previously found no significant difference in the risk of heart failure by race overall (14); however, in the present study, when we took into account region and sex, we found a marginally lower risk of heart failure in southern black men and women than in their white counterparts. Our failure in the present study (except for women in other regions) to find a greater risk of CVA in blacks is consistent with one prior report (11).

We are not aware of other reports examining differences in CVD outcomes among American populations with diabetes that have demonstrated variation by race and region. Although we found a protective effect of southern residence for blacks and a detrimental effect of non-southern residence for black women, these findings should be accepted with caution. The apparent protective effect of black race in the South may be due to survivor bias if southern blacks with diabetes tend to develop CVD and die before becoming eligible for Medicare. Alternatively, there may be racial variation between the South and elsewhere in terms of the diagnosis of AMI or IHD or at least in the choice of ICD-9 codes. Differential CVD risk by region may also reflect differences in unmeasured CVD risk factors and health behaviors such as smoking, diet, control of diabetes, hypertension, or dyslipidemia. Our findings are not likely due to systematically better control of these risk factors in blacks versus whites, since the preponderance of evidence suggests blacks have both worse quality of diabetes care and control of risk factors for CVD (58,21).

The chief strengths of this analysis are the use of a large, longitudinal dataset with a population representative of older Americans with diabetes, inclusion of geographic region into the study, and the use of regression models incorporating several variables to address differences in risk of CVD. The dataset allowed us to precisely quantify even small differences in risk between subgroups defined by sex, region, and race. The main limitation is the use of claims data for all aspects of this work, including the determination of outcomes as well as the covariates. Furthermore, age, sex, and race were not independently verified. These limitations raise concerns regarding misclassification. Our criteria for diabetes using Medicare claims were reported to have a specificity of 98% (26). High specificity for many diagnoses including IHD and CVA has been reported when comparing claims to clinical records (27). In addition, while we excluded those with a 1994 claim for CVD in incidence analyses, we could not ascertain which individuals had had a CVD claim before 1994. The most likely effects of the limitations are to make less precise the incidence data and an obscuring of associations between predictor variables and the outcomes of interest, which would likely cause us to underestimate the true effects of these variables. Another limitation is our inability to evaluate additional factors related to cardiovascular risk, including the duration, severity, or control of diabetes, smoking, or hyperlipidemia.

How best to explain the greater risk for CVD of black women with diabetes not residing in the South (versus white women with diabetes) is open to question. The finding of a regional effect argues against any genetic explanation and suggests that regional differences in lifestyle, environment, or quality of care may play some role. We note, however, that among the elderly with diabetes, CVD occurs frequently regardless of race, sex, or region. We also point out that while many comparisons reach statistical significance, this was often abetted by large samples, and the differences observed are quite small in magnitude. The high incidence of CVD demonstrated among the elderly with diabetes is especially concerning given the recent trends in diabetes incidence in the elderly (28). The aging of the U.S. population and the high risk of diabetes in older people make prevention of diabetes and its complications a high priority.

Figure 1—

Kaplan-Meier survival free of ischemic heart disease (IHD) by race, sex, and region among elderly black and white Medicare beneficiaries with diabetes from January 1995 to December 1999. Other regions = northeast, midwest, and west.

Figure 1—

Kaplan-Meier survival free of ischemic heart disease (IHD) by race, sex, and region among elderly black and white Medicare beneficiaries with diabetes from January 1995 to December 1999. Other regions = northeast, midwest, and west.

Close modal
Table 1—

Selected characteristics of Medicare beneficiaries >65 years old with diabetes in 1994 by race

WhiteBlack
n 126,153 (100) 17,962 (100) 
Sex   
    Women 73,783 (58.5) 12,614 (70.2) 
    Men 52,370 (41.5) 5,348 (29.8) 
Age-group (years)   
    65–69 33,981 (26.9) 5,674 (31.6) 
    70–74 35,685 (28.3) 5,144 (28.6) 
    75–79 27,990 (22.2) 3,601 (20.0) 
    80–84 17,455 (13.8) 2,139 (11.9) 
    >85 11,042 (8.8) 1,404 (7.8) 
Region   
    Northeast 32,621 (25.9) 2,939 (16.4) 
    South 41,895 (33.2) 10,617 (59.1) 
    Midwest 35,502 (28.1) 3,541 (19.7) 
    West 16,135 (12.8) 865 (4.8) 
End-stage renal disease 758 (0.6) 409 (2.3) 
Nephropathy 7,708 (6.1) 1,914 (10.7) 
Hypertension 78,462 (62.2) 13,574 (75.6) 
Neuropathy* 19,413 (15.4) 2,662 (14.8) 
Retinopathy 18,973 (15.0) 3,115 (17.3) 
AMI 6,455 (5.1) 734 (4.1) 
IHD 50,355 (39.9) 5,611 (31.2) 
Heart failure 28,290 (22.4) 4,052 (22.6) 
CVA* 23,039 (18.3) 3,405 (19.0) 
WhiteBlack
n 126,153 (100) 17,962 (100) 
Sex   
    Women 73,783 (58.5) 12,614 (70.2) 
    Men 52,370 (41.5) 5,348 (29.8) 
Age-group (years)   
    65–69 33,981 (26.9) 5,674 (31.6) 
    70–74 35,685 (28.3) 5,144 (28.6) 
    75–79 27,990 (22.2) 3,601 (20.0) 
    80–84 17,455 (13.8) 2,139 (11.9) 
    >85 11,042 (8.8) 1,404 (7.8) 
Region   
    Northeast 32,621 (25.9) 2,939 (16.4) 
    South 41,895 (33.2) 10,617 (59.1) 
    Midwest 35,502 (28.1) 3,541 (19.7) 
    West 16,135 (12.8) 865 (4.8) 
End-stage renal disease 758 (0.6) 409 (2.3) 
Nephropathy 7,708 (6.1) 1,914 (10.7) 
Hypertension 78,462 (62.2) 13,574 (75.6) 
Neuropathy* 19,413 (15.4) 2,662 (14.8) 
Retinopathy 18,973 (15.0) 3,115 (17.3) 
AMI 6,455 (5.1) 734 (4.1) 
IHD 50,355 (39.9) 5,611 (31.2) 
Heart failure 28,290 (22.4) 4,052 (22.6) 
CVA* 23,039 (18.3) 3,405 (19.0) 

Data are n(%). All differences are statistically significant at P < 0.001, except for

*

P < 0.05 and

not significantly different.

Table 2—

Incidence of CVD (specific types or any) in 1995–1999 among Medicare beneficiaries with diabetes who were without disease in 1994

ExcludedNumber at riskIncident casesIncidence per 100 person-years95% CI
AMI (south)      
    White women 986 23,381 3,460 4.1 4.0–4.3 
    Black women 248 7,427 931 3.4 3.2–3.7 
    White men 984 16,544 2,906 5.2 5.0–5.4 
    Black men 115 2,827 357 3.8 3.4–4.2 
AMI (other region)      
    White women 2358 47,058 7,465 4.6 4.5–4.7 
    Black women 238 4,701 739 4.6 4.3–5.0 
    White men 2,127 32,715 5,968 5.5 5.4–5.6 
    Black men 133 2,273 346 4.7 4.3–5.3 
IHD (south)      
    White women 8,568 15,799 7,280 15.9 15.5–16.2 
    Black women 1,986 5,689 2,333 13.7 13.2–14.3 
    White men 7,599 9,929 4,884 18.3 17.8–18.8 
    Black men 850 2,092 822 14.1 13.1–15.1 
IHD (other region)      
    White women 18,595 30,821 14,485 17.0 16.7–17.3 
    Black women 1,836 3,103 1,565 19.7 18.7–20.7 
    White men 15,593 19,249 9,736 19.6 19.2–20.0 
    Black men 939 1,467 699 19.3 17.9–20.8 
Heart failure (south)      
    White women 5,529 18,838 7,671 12.8 12.5–13.0 
    Black women 1,734 5,941 2,334 12.3 11.5–13.1 
    White men 3,593 13,935 5,409 12.4 12.1–12.8 
    Black men 585 2,357 859 12.5 11.9–13.0 
Heart failure (other region)      
    White women 11,540 37,876 15,403 13.2 12.9–13.4 
    Black women 1,734 3,745 1,561 14.2 13.5–15.0 
    White men 3,593 27,214 10,824 13.1 12.8–13.3 
    Black men 585 1,867 735 12.4 12.1–12.8 
CVA (south)      
    White women 4,646 19,721 8,083 13.5 13.2–13.8 
    Black Women 1,375 6,300 2,441 12.3 11.7–12.8 
    White men 3,459 14,069 5,504 13.3 12.9–13.6 
    Black men 557 2,385 914 13.0 12.9–13.6 
CVA (other regions)      
    White women 8,633 40,783 15,223 12.3 12.1–12.5 
    Black women 1,017 3,922 1,627 14.5 13.8–15.2 
    White men 6,301 28,541 10,667 12.8 12.6–13.1 
    Black men 456 1,950 712 12.8 11.9–13.8 
Any CVD (south)*      
    White women 12,369 11,998 7,768 25.5 24.9–26.0 
    Black women 3,462 4,213 2,635 24.3 22.4–25.5 
    White men 9,760 7,768 4,976 26.1 25.4–26.9 
    Black men 1,358 1,584 940 23.9 23.4–25.2 
Any CVD (other regions)      
    White women 25,476 23,940 15,093 25.7 25.3–26.1 
    Black women 2,589 2,350 1,533 29.2 27.7–30.7 
    White men 19,593 15,249 9,900 27.6 27.1–28.2 
    Black men 1,297 1,109 671 27.0 25.0–29.1 
ExcludedNumber at riskIncident casesIncidence per 100 person-years95% CI
AMI (south)      
    White women 986 23,381 3,460 4.1 4.0–4.3 
    Black women 248 7,427 931 3.4 3.2–3.7 
    White men 984 16,544 2,906 5.2 5.0–5.4 
    Black men 115 2,827 357 3.8 3.4–4.2 
AMI (other region)      
    White women 2358 47,058 7,465 4.6 4.5–4.7 
    Black women 238 4,701 739 4.6 4.3–5.0 
    White men 2,127 32,715 5,968 5.5 5.4–5.6 
    Black men 133 2,273 346 4.7 4.3–5.3 
IHD (south)      
    White women 8,568 15,799 7,280 15.9 15.5–16.2 
    Black women 1,986 5,689 2,333 13.7 13.2–14.3 
    White men 7,599 9,929 4,884 18.3 17.8–18.8 
    Black men 850 2,092 822 14.1 13.1–15.1 
IHD (other region)      
    White women 18,595 30,821 14,485 17.0 16.7–17.3 
    Black women 1,836 3,103 1,565 19.7 18.7–20.7 
    White men 15,593 19,249 9,736 19.6 19.2–20.0 
    Black men 939 1,467 699 19.3 17.9–20.8 
Heart failure (south)      
    White women 5,529 18,838 7,671 12.8 12.5–13.0 
    Black women 1,734 5,941 2,334 12.3 11.5–13.1 
    White men 3,593 13,935 5,409 12.4 12.1–12.8 
    Black men 585 2,357 859 12.5 11.9–13.0 
Heart failure (other region)      
    White women 11,540 37,876 15,403 13.2 12.9–13.4 
    Black women 1,734 3,745 1,561 14.2 13.5–15.0 
    White men 3,593 27,214 10,824 13.1 12.8–13.3 
    Black men 585 1,867 735 12.4 12.1–12.8 
CVA (south)      
    White women 4,646 19,721 8,083 13.5 13.2–13.8 
    Black Women 1,375 6,300 2,441 12.3 11.7–12.8 
    White men 3,459 14,069 5,504 13.3 12.9–13.6 
    Black men 557 2,385 914 13.0 12.9–13.6 
CVA (other regions)      
    White women 8,633 40,783 15,223 12.3 12.1–12.5 
    Black women 1,017 3,922 1,627 14.5 13.8–15.2 
    White men 6,301 28,541 10,667 12.8 12.6–13.1 
    Black men 456 1,950 712 12.8 11.9–13.8 
Any CVD (south)*      
    White women 12,369 11,998 7,768 25.5 24.9–26.0 
    Black women 3,462 4,213 2,635 24.3 22.4–25.5 
    White men 9,760 7,768 4,976 26.1 25.4–26.9 
    Black men 1,358 1,584 940 23.9 23.4–25.2 
Any CVD (other regions)      
    White women 25,476 23,940 15,093 25.7 25.3–26.1 
    Black women 2,589 2,350 1,533 29.2 27.7–30.7 
    White men 19,593 15,249 9,900 27.6 27.1–28.2 
    Black men 1,297 1,109 671 27.0 25.0–29.1 
*

Incidence of any CVD(AMI, CVA, IHD, heart failure) outcome among individuals with diabetes without any CVD in 1994.

Table 3—

Association between selected risk factors and incidence of CVD among individuals with diabetes without specified disease in 1994

Risk factorAMIIHDHeart failureCVAAny CVD
End-stage renal disease 1.35 (1.16–1.56) 1.71 (1.51–1.93) 2.02 (1.79–2.27) 1.42 (1.28–1.58) 1.95 (1.69–2.26) 
Renal disease 1.52 (1.44–1.60) 1.41 (1.34–1.47) 1.67 (1.60–1.74) 1.42 (1.36–1.48) 1.43 (1.36–1.51) 
Hypertension 1.22 (1.18–1.25) 1.15 (1.12–1.17) 1.17 (1.14–1.19) 1.13 (1.10–1.15) 1.10 (1.08–1.12) 
Neuropathy 1.22 (1.18–1.26) 1.22 (1.19–1.26) 1.22 (1.19–1.26) 1.25 (1.22–1.28) 1.22 (1.19–1.25) 
Retinopathy 1.30 (1.25–1.34) 1.18 (1.15–1.21) 1.26 (1.22–1.29) 1.21 (1.18–1.24) 1.20 (1.17–1.23) 
Sex (men versus women)      
    White 1.28 (1.24–1.32) 1.18 (1.16–1.21) 1.06 (1.04–1.09) 1.08 (1.06–1.10) 1.10 (1.08–1.13) 
    Black 1.10 (1.01–1.20) 1.04 (0.98–1.10) 1.04 (0.99–1.11) 1.03 (0.97–1.09) 1.01 (0.95–1.07) 
Region (south versus other)      
    White 0.94 (0.91–0.96) 0.95 (0.93–0.97) 0.99 (0.96–1.01) 1.09 (1.07–1.12) 1.00 (0.98–1.02) 
    Black 0.77 (0.71–0.84) 0.72 (0.68–0.76) 0.88 (0.83–0.93) 0.91 (0.86–0.96) 0.86 (0.81–0.90) 
Race (black versus white)      
    South      
Women 0.80 (0.75–0.86) 0.84 (0.81–0.88) 0.95 (0.91–0.99) 0.92 (0.89–0.96) 0.95 (0.91–0.99) 
Men 0.69 (0.63–0.75) 0.74 (0.69–0.78) 0.93 (0.88–0.99) 0.88 (0.83–0.93) 0.87 (0.82–0.92) 
    Other regions      
Women 0.97 (0.90–1.04) 1.11 (1.06–1.16) 1.06 (1.01–1.11) 1.12 (1.06–1.17) 1.10 (1.05–1.16) 
Men 0.83 (0.76–0.91) 0.97 (0.91–1.03) 1.04 (0.98–1.11) 1.06 (1.00–1.12) 1.01 (0.95–1.07) 
Risk factorAMIIHDHeart failureCVAAny CVD
End-stage renal disease 1.35 (1.16–1.56) 1.71 (1.51–1.93) 2.02 (1.79–2.27) 1.42 (1.28–1.58) 1.95 (1.69–2.26) 
Renal disease 1.52 (1.44–1.60) 1.41 (1.34–1.47) 1.67 (1.60–1.74) 1.42 (1.36–1.48) 1.43 (1.36–1.51) 
Hypertension 1.22 (1.18–1.25) 1.15 (1.12–1.17) 1.17 (1.14–1.19) 1.13 (1.10–1.15) 1.10 (1.08–1.12) 
Neuropathy 1.22 (1.18–1.26) 1.22 (1.19–1.26) 1.22 (1.19–1.26) 1.25 (1.22–1.28) 1.22 (1.19–1.25) 
Retinopathy 1.30 (1.25–1.34) 1.18 (1.15–1.21) 1.26 (1.22–1.29) 1.21 (1.18–1.24) 1.20 (1.17–1.23) 
Sex (men versus women)      
    White 1.28 (1.24–1.32) 1.18 (1.16–1.21) 1.06 (1.04–1.09) 1.08 (1.06–1.10) 1.10 (1.08–1.13) 
    Black 1.10 (1.01–1.20) 1.04 (0.98–1.10) 1.04 (0.99–1.11) 1.03 (0.97–1.09) 1.01 (0.95–1.07) 
Region (south versus other)      
    White 0.94 (0.91–0.96) 0.95 (0.93–0.97) 0.99 (0.96–1.01) 1.09 (1.07–1.12) 1.00 (0.98–1.02) 
    Black 0.77 (0.71–0.84) 0.72 (0.68–0.76) 0.88 (0.83–0.93) 0.91 (0.86–0.96) 0.86 (0.81–0.90) 
Race (black versus white)      
    South      
Women 0.80 (0.75–0.86) 0.84 (0.81–0.88) 0.95 (0.91–0.99) 0.92 (0.89–0.96) 0.95 (0.91–0.99) 
Men 0.69 (0.63–0.75) 0.74 (0.69–0.78) 0.93 (0.88–0.99) 0.88 (0.83–0.93) 0.87 (0.82–0.92) 
    Other regions      
Women 0.97 (0.90–1.04) 1.11 (1.06–1.16) 1.06 (1.01–1.11) 1.12 (1.06–1.17) 1.10 (1.05–1.16) 
Men 0.83 (0.76–0.91) 0.97 (0.91–1.03) 1.04 (0.98–1.11) 1.06 (1.00–1.12) 1.01 (0.95–1.07) 

Data are hazard ratio(95% CI) from a multivariate model simultaneously adjusting for comorbidities listed, age, sex, race, and region. Reference category for comorbidities is individuals without the comorbidity.

This publication was made possible through a cooperative agreement between the Centers for Disease Control and Prevention (CDC) and the Association of Teachers of Preventive Medicine (ATPM), award number TS-0778.

Its contents are the responsibility of the authors and do not necessarily reflect the official views of the CDC or ATPM.

1.
Centers for Disease Control and Prevention: Trends in ischemic heart disease death rates for blacks and whites: U.S., 1981–1995.
MMWR Morb Mortal Wkly Rep
47
:
945
–949,
1998
2.
Howard G, Howard VJ: Ethnic disparities in stroke: the scope of the problem.
Ethn Dis
11
:
761
–768,
2001
3.
Gaines K, Burke G: Ethnic differences in stroke: black-white differences in the United States population: SECORDS Investigators.
Neuroepidemiology
14
:
209
–239,
1995
4.
Kannel WB, McGee DL: Diabetes and cardiovascular disease: the Framingham study.
JAMA
241
:
2035
–2038,
1979
5.
Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE, Little RR, Wiedmeyer HM, Byrd-Holt DD: 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
21
:
518
–524,
1998
6.
Beckles GL, Engelgau MM, Narayan KM, Herman WH, Aubert RE, Williamson DF: Population-based assessment of the level of care among adults with diabetes in the U.S.
Diabetes Care
21
:
1432
–1438,
1998
7.
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
136
:
565
–574,
2002
8.
Kirk JK, Bell RA, Bertoni AG, Arcury TA, Quandt SA, Goff DC Jr, Narayan KM: Qualitative review of studies of diabetes preventive care among minority patients in the United States, 1993–2003 (Review)
Am J Manag Care
11
:
349
–360,
2005
9.
Ness J, Aronow WS: Prevalence of coronary artery disease, ischemic stroke, peripheral arterial disease, and coronary revascularization in older African-Americans, Asians, Hispanics, whites, men, and women.
Am J Cardiol
84
:
932
–933, A7,
1999
10.
Gillum RF, Mussolino ME, Madans JH: Diabetes mellitus, coronary heart disease incidence, and death from all causes in African American and European American women: the NHANES I epidemiologic follow-up study.
J Clin Epidemiol
53
:
511
–518,
2000
11.
Karter AJ, Ferrara A, Liu JY, Moffet HH, Ackerson LM, Selby JV: Ethnic disparities in diabetic complications in an insured population.
JAMA
287
:
2519
–2527,
2002
12.
Young BA, Maynard C, Boyko EJ: Racial differences in diabetic nephropathy, cardiovascular disease, and mortality in a national population of veterans.
Diabetes Care
26
:
2392
–2399,
2003
13.
Chaturvedi N, Jarrett J, Morrish N, Keen H, Fuller JH: Differences in mortality and morbidity in African Caribbean and European people with non-insulin dependent diabetes mellitus.
BMJ
313
:
848
–852,
1996
14.
Bertoni AG, Hundley WG, Massing MW, Bonds DE, Burke GL, Goff DC, Jr: Heart failure prevalence, incidence, and mortality in the elderly with diabetes.
Diabetes Care
27
:
699
–703,
2004
15.
American Heart Association:
Heart Disease and Stroke Statistics: 2004 Update
. Dallas, TX, American Heart Association,
2003
16.
Pickle LW, Mungiole M, Jones GK, White AA:
Atlas of United States Mortality
. Hyattsville, MD, National Center for Health Statistics,
1996
17.
Mokdad AH, Bowman BA, Ford ES, Vinicor F, Marks JS, Koplan JP: The continuing epidemics of obesity and diabetes in the United States.
JAMA
286
:
1195
–1200,
2001
18.
Krop JS, Powe NR, Weller WE, Shaffer TJ, Saudek CD, Anderson GF: Patterns of expenditures and use of services among older adults with diabetes: implications for the transition to capitated managed care.
Diabetes Care
21
:
747
–752,
1998
19.
Census Bureau:
National Population Estimates 1990–1999
[article online]. Available from http://www.census.gov. Accessed 16 February
2005
20.
Korn EL, Graubard BI, Midthune D: Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale.
Am J Epidemiol
145
:
72
–80,
1997
21.
Chin MH, Zhang JX, Merrell K: Diabetes in the African-American Medicare population: morbidity, quality of care, and resource utilization.
Diabetes Care
21
:
1090
–1095,
1998
22.
Adler AI, Stratton IM, Neil HA, Yudkin JS, Matthews DR, Cull CA, Wright AD, Turner RC, Holman RR: Association of systolic blood pressure with macrovascular and microvascular complications of type 2 diabetes (UKPDS 36): prospective observational study.
BMJ
321
:
412
–419,
2000
23.
Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA, Hadden D, Turner RC, Holman RR: Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study.
BMJ
321
:
405
–412,
2000
24.
DeStefano F, Newman J: Comparison of coronary heart disease mortality risk between black and white people with diabetes.
Ethn Dis
3
:
145
–151,
1993
25.
Centers for Disease Control and Prevention:
Diabetes Surveillance, 1999
. Atlanta, GA, U.S. Department of Health and Human Services,
1999
26.
Hebert PL, Geiss LS, Tierney EF, Engelgau MM, Yawn BP, McBean AM: Identifying persons with diabetes using medicare claims data.
Am J of Medical Quality
14
:
270
–277,
1999
27.
Jollis JG, Ancukiewicz M, DeLong ER, Pryor DB, Muhlbaier LH, Mark DB: Discordance of databases designed for claims payment versus clinical information systems: implications for outcomes research
Ann Intern Med
119
:
844
–850,
1993
28.
Narayan KM, Boyle JP, Thompson TJ, Sorensen SW, Williamson DF: Lifetime risk for diabetes mellitus in the United States.
JAMA
290
:
1884
–1890,
2003

J.K.K. has received honoraria/consulting fees from Novo Nordisk and grant/research support from Novartis and AstraZeneca.

Additional information for this article can be found in an online appendix at http://care.diabetesjournals.org.

A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.