To our knowledge, this is the first study to obtain estimates based on clinical data of CHD, stroke, and PVD for the six largest Asian subpopulations (Asian Indians, Chinese, Filipino, Japanese, Korean, Vietnamese) in the United States. These findings support previous research that has suggested considerable heterogeneity in Asian-American subgroups with regard to CHD, stroke, and PVD prevalence. Our findings indicate that Filipinos and Asian Indians generally diverge from the other Asian subgroups, with both groups showing elevated risk for CHD and Filipino women elevated risk for overall stroke. When subclasses of stroke were examined, significantly greater risk of ischemic stroke was found for Filipino women, and greater risk of hemorrhagic stroke was found for Vietnamese men and Korean women compared with NHWs. Significantly lower risk for CHD was found for Chinese men and women compared with NHWs. Because some of the subgroups have greater rates than NHWs and some lower, the ORs comparing the aggregate Asian group to NHWs for CHD and stroke are close to one and not statistically significant. In contrast, odds of PVD were generally lower among all Asian subgroups and the aggregate Asian group compared with NHWs.
The majority of studies in which the authors examined CVD in Asian Americans have focused on one Asian subgroup alone. The Ni-Hon-San study, a multisite study that included the Honolulu Heart Study, has tracked CVD in Japanese men since 1965 and demonstrated greater rates of CHD and lower rates of stroke for 16,580 Japanese Americans compared with 9,329 native Japanese (20
). The MESA study included 797 Chinese Americans and demonstrated lower congestive heart failure incidence and PVD prevalence rates for this subgroup compared with white subjects (15
). With only one subgroup in each of these studies, it is difficult to compare findings across Asian subgroups.
Klatsky and Tekawa used hospital record data collected approximately 30 years ago (1978–1985) from California to examine several Asian-American subgroups, including Chinese, Filipinos, Japanese, and Asian Indians. Although the Asian-American population has changed dramatically since, Klatsky and Tekawa similarly found that Asian Indian men and Filipino women were at greater risk whereas Chinese were at lower risk of hospitalization for CHD (12
). Additional studies that used the same database found lower risk of PVD for Chinese Americans (12
), Filipino Americans (12
), and aggregated Asians (12
) compared with NHWs. The odds of CHD, stroke, and PVD among the Asian subgroups in our study population were generally similar to those reported by Klatsky and Tekawa. The studies by Klatsky and Tekawa were limited to hospitalization data, which would fail to capture patients that do not seek immediate medical care or receive care only in the outpatient setting. Our data source provides more recent data (2007–2009), two additional subgroups, and better captures patients ever diagnosed with CHD, stroke, and PVD. Nonetheless the consistency of our findings suggests the contrasts we found are not unique to this study population.
The only nationally representative data for CVD in Asian Americans is the NHIS, which is limited by self report and small sample sizes in the Asian subgroups. Our study sample is not national, and therefore our point estimates are not generalizable. However, relative differences in prevalence rates between the Asian subgroups are likely to be similar across the country, and thus help to identify highrisk groups. Our rates of CHD for NHWs and Asian Americans were generally lower than those reported by NHIS. NHIS did not report stroke rates for some of the subgroups. Although our rates of stroke were lower for NHWs and Chinese, compared with those of NHIS, the decrease was similar for both groups. Differences between our prevalence rates and those reported by NHIS may be due to differences in measurement (self-report versus health records) or may be to the result of real differences between our study population and the general U.S. population, reflecting geography, education, economic status or access to care.
Differences in traditional CVD risk factors between Asian-American subgroups may contribute to the observed differences in our CVD prevalence rates. Filipinos and Asian Indians have been shown to have greater prevalence rates of some traditional risk factors for CHD (24
). Higher prevalence rates have been reported for type 2 diabetes among Filipinos (25
), and for type 2 diabetes (27
) and physical inactivity (24
) among Asian Indians compared with NHWs. The biggest risk factor for stroke is hypertension, which is a factor in nearly 70% of all strokes (30
). Greater rates of hypertension have been reported for Filipinos compared to NHWs and other Asian-American subgroups (9
). Few investigators have examined PVD prevalence and risk factors among Asian Americans. Authors of the MESA study found that after adjusting for traditional (age, smoking, diabetes, body mass index, hypertension, and dyslipidemia) and novel risk factors, the OR for peripheral arterial disease was significantly lower for Chinese subjects, suggesting that intrinsic or unrecognized factors associated with race/ethnicity may account for this reduced risk (15
). It has been hypothesized that genetic factors, particularly those involving clotting, may play a role in the reduced risk of PVD in Asian subjects (15
). We found that most Asian-American subgroups had lower prevalence of arterial and venous disease, compared to NHWs.
Limitations of this study include a single geographic area, with somewhat limited sample size in the smaller Asian subgroups (ie, Korean and Vietnamese populations). Although the Ni-Hon-San study demonstrates the importance of separating U.S-born from foreign-born, we were unable to identify migration status in our patient population. However, only 3% of our patients are limited English-proficient speakers, which may indicate a high level of acculturation among the Asian patient population. The small number of cases of CHD, stroke, and PVD, limit our ability to find significant differences for those smaller groups despite suggestive point estimates. CVD prevalence definitions were based on outpatient visits only. Although this may lead to underestimation of CVD prevalence, it is likely in this group of insured patients that any hospitalizations are followed by a follow-up outpatient visit with one of our providers. While, the exclusion of transient cerebral ischemia and other/ill-defined and late effects of cerebrovascular disease may underestimate actual cases of stroke, the specificity of overall stroke, ischemic and hemorrhagic stroke is improved. The authors of previous studies have validated administrative data in identifying diagnoses of CHD and stroke and have shown them to be particularly useful in making high-level comparisons (31
). The relative homogeneity in economic status in this clinical population (all insured, with access to health care) improves the internal validity of our comparisons.
In conclusion, our findings show marked heterogeneity in CVD prevalence among Asian-American subgroups, which is obscured when Asian subgroups are aggregated. Filipinos and Asian Indians appear to be at greater risk for CHD, and Filipino women at greater risk for stroke compared with NHWs and other Asian subgroups. Targeted prevention and treatment efforts may be especially needed for Filipinos and Asian Indians. Data on CVD risk factors and outcomes in Vietnamese and Korean American subpopulations are lacking and may underestimate risk in these subgroups. More research, particularly with a nationally representative population-based sample, needs to investigate CVD risk factors and outcomes separately among Asian subgroups to better understand variation in these disease patterns. The American Heart Association recently issued a scientific advisory (32
) recommending changes in data collection—such as disaggregation of Asian Americans in the myocardial infarction (4
) and stroke (5
) registries, development of better standard measurements of diet and acculturation, and new research studies to improve health disparities among Asian Americans. These critical data are needed to better inform preventive and treatment interventions for CVD among Asian Americans.