To examine sex and racial/ethnic differences in cardiovascular risk factor treatment and control among individuals with diabetes in the Multi-Ethnic Study of Atherosclerosis (MESA).
This study was an observational study examining mean levels of cardiovascular risk factors and proportion of subjects achieving treatment goals.
The sample included 926 individuals with diabetes. Compared with men, women were 9% less likely to achieve LDL cholesterol <130 mg/dl (adjusted prevalence ratio 0.91 [0.83–0.99]) and systolic blood pressure (SBP) <130 mmHg (adjusted prevalence ratio 0.91 [0.85–0.98]). These differences diminished over time. A lower percentage of women used aspirin (23 vs. 33%; P < 0.001). African American and Hispanic women had higher mean levels of SBP and lower prevalence of aspirin use than non-Hispanic white women.
Women with diabetes had unfavorable cardiovascular risk factor profiles compared with men. African American and Hispanic women had less favorable profiles than non-Hispanic white women.
Population-based health survey data suggest that sex and racial/ethnic disparities are present in diabetes process of care measures and cardiovascular risk factor control (1,,,,,,,–9). Available data also indicate that sex-specific race/ethnicity differences are present in cardiovascular risk factor control, but these data are limited to Medicare and Veterans' Hospital patient populations (5,10,,–13). We therefore performed analyses of participants with diabetes in the Multi-Ethnic Study of Atherosclerosis (MESA) to examine sex and sex-specific racial/ethnic differences in cardiovascular risk factor treatment and control.
RESEARCH DESIGN AND METHODS
MESA is a multicenter cohort study of 6,814 men and women age 45–84 years with no clinical evidence of cardiovascular disease at time of enrollment (14). Four MESA exam periods occurred between 2000 and 2007. The study was approved by the Institutional Review Board at all participating institutions, and all MESA participants provided informed consent. Criteria for recruitment, data collection methods, and laboratory techniques have been previously described (14). Participants were classified as having diabetes if at exam 1 they had a fasting plasma glucose ≥126 mg/dl, used oral hypoglycemic agents and/or insulin, or reported a physician diagnosis of diabetes.
Multivariate models were used to calculate predicted means for lipid, blood pressure, and A1C levels. Prevalence ratios (PRs) were calculated for percentages of participants achieving cardiovascular risk factor goals using binomial regression, with adjustment for age, MESA site, socioeconomic status variables, and either sex or race/ethnicity. Potential effects of selective attrition on longitudinal results were examined using t tests to determine if participants lost to follow-up had higher mean cardiovascular risk factor levels compared with those retained in the study. Analyses were performed using SAS version 9.1.
Of 926 MESA participants with diabetes at exam 1 (2000–2002), 48% were women. Four racial/ethnic groups were represented (19% non-Hispanic white [NHW], 38% African American, 31% Hispanic, and 12% Chinese). Compared with men, women were more likely to report gross family income <$20,000 (42 vs. 26%) and less than high school education (33 vs. 27%).
At exam 1, after adjustment for age, MESA site, and race/ethnicity, systolic blood pressure (SBP) was 3.5 mmHg higher among women than men (133.7 vs. 130.2 mmHg, P < 0.01). LDL cholesterol and A1C did not differ by sex. After additional adjustment for socioeconomic status, there was no sex difference in mean SBP (Table 1 ). African American and Hispanic women had significantly higher mean SBP values than NHW women. A significantly lower percentage of women were taking aspirin (Table 1). Hispanic women reported taking less aspirin than NHW women.
A lower percentage of women achieved LDL cholesterol <130 mg/dl (69.4 vs. 77.1%, P = 0.01) and SBP <130 mmHg (42.2 vs. 57.8%, P = 0.002) compared with men. Women were 9% less likely to achieve LDL cholesterol <130 mg/dl than men (adjusted PR 0.91 [0.85–0.98]) and 9% less likely to achieve SBP <130 mmHg (adjusted PR 0.91 [0.83–0.99]). African American and Hispanic women were 31% (adjusted PR 0.69 [0.51–0.91]) and 30% (adjusted PR 0.70 [0.52–0.95]) less likely, respectively, to achieve blood pressure <130/80 mmHg compared with NHW women.
Of 926 subjects with diabetes at exam 1, 802 completed exam 2, 751 exam 3, and 719 exam 4. At exam 4, LDL cholesterol (96.4 vs. 94.6 mg/dl, P = 0.54) and SBP (130.3 vs. 127.6 mmHg, P = 0.11) did not differ between women and men. At exam 4, there was no difference in the percentage of women achieving LDL cholesterol <130 mg/dl (adjusted PR 0.97 [0.91–1.04]) or SBP <130 mmHg (adjusted PR 0.95 [0.88–1.02]) compared with men. African American women were 33% less likely to achieve blood pressure <130/80 mmHg (adjusted PR 0.67 [0.49–0.91]) compared with NHW women. Although women reported higher antihypertensive and lipid-lowering medication use compared with men at exam 1, there was no difference at exam 4 in antihypertensive (80 vs. 75%, P = 0.20) or lipid-lowering medication (51 vs. 49%, P = 0.70) use. Aspirin use increased for both sexes from 2000 to 2007; however, women remained less likely to report aspirin use at exam 4 compared with men (44 vs. 57%, P < 0.05). Aspirin use remained lower for African American and Hispanic women compared with NHW women (51 and 39% vs. 58%, respectively).
Among MESA participants with diabetes, at the baseline exam, a lower proportion of women achieved consensus treatment targets for SBP and LDL cholesterol, after adjustment for covariates. These differences were observed despite a greater reported use of blood pressure and lipid-lowering medications among women. African American and Hispanic women had lower proportions achieving consensus treatment goals for blood pressure and less reported use of aspirin compared with NHW women. The sex difference in LDL cholesterol and SBP control diminished over time.
Our findings are consistent with previous reports showing sex and racial/ethnic differences in cardiovascular risk factor control among individuals with diabetes (2,12,13,15). Inconsistent findings have been reported regarding sex and racial/ethnic differences in use of medications for management of cardiovascular factors (3,4,8).
There are several potential mechanisms for the sex and racial/ethnic differences we observed. Regression adjustment for socioeconomic status variables eliminated the significant sex difference in mean SBP. Socioeconomic variables may correlate with access to high-quality medical care and with personal behaviors that influence risk factor levels, including medication adherence, diet, and exercise. Our findings may also indicate a disparity in medication titration by physicians.
There are several limitations to our study. Information on medication use was self-reported and patient adherence was not recorded. Study participation may also have influenced treatment patterns because exam results were reported to the participants. Despite these limitations, our findings provide new information about sex and sex-specific race/ethnicity differences in cardiovascular risk factor treatment and control among individuals with diabetes in a contemporary multiethnic cohort with diverse sources of insurance.
In conclusion, we found that among subjects with diabetes in MESA, women had unfavorable cardiovascular risk factor profiles compared with men at baseline exam; however, these differences diminished over time. African American and Hispanic women had less favorable profiles than NHW women.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This research was supported by the National Heart, Lung, and Blood Institute (NHLBI) contracts N01-HC-95159 through N01-HC-95169 and General Medicine Training Grant D55HP05154-01-00 (Health Services Research Administration [HRSA]). G.J.W. has received support from the HRSA through a postdoctoral research training grant. R.G.B. has received support from NHLBI and the Robert Wood Johnson Foundation. O.C. has received support from HRSA, NHLBI, and the National Institute on Aging (NIA). A.G.B. has received support from NHLBI and the National Institute of Diabetes and Digestive and Kidney Diseases. S.S. has received support from NHLBI, NIA, HRSA, the Centers for Medicare and Medicaid Services (CMS), and the American Cancer Society.
No potential conflicts of interest relevant to this article were reported.