Author + information
- Received December 2, 2013
- Revision received February 4, 2014
- Accepted February 8, 2014
- Published online August 1, 2014.
- Diana Choi, PhD∗,
- Edson Nemi, MD†,
- Carlos Fernando, MD†,
- Milan Gupta, MD†,‡,§ and
- Gordon W. Moe, MD†,§∗ ()
- ∗Faculty of Medicine, University of Toronto, Toronto, Ontario, Canada
- †Division of Cardiology, Department of Medicine, St Michael’s Hospital, University of Toronto, Toronto, Ontario, Canada
- ‡Division of Cardiology, McMaster University, Hamilton, Ontario, Canada
- §Keenan Research Center in the Li Ka Shing Knowledge Institute, St Michael's Hospital, Toronto, Ontario, Canada
- ↵∗Reprint requests and correspondence:
Dr. Gordon W. Moe, Division of Cardiology, Department of Medicine, St Michael’s Hospital, 30 Bond Street, Toronto, Ontario M5B 1W8, Canada.
Objectives The purpose of our study was to compare the clinical features of Chinese and South Asians, the 2 largest minority populations in Canada, with non-Chinese/non-South Asian (NCH/NSA) patients managed in common social macroenvironments and healthcare systems.
Background Heart failure is an increasingly prevalent condition. Although ethnic minorities comprise a growing proportion of the population in Western countries, the clinical profiles of ethnic minorities with heart failure are largely unknown.
Methods We analyzed records of patients with heart failure managed in 2 specialized clinics in Ontario, Canada. Of the 1,671 patients, 181 (11%) were identified as Chinese and 215 (13%) as South Asian.
Results Our analyses showed that fewer Chinese patients were found to have a history of myocardial infarction (MI) (30% vs. 52%), 3 occluded/stenosed coronary vessels on angiogram (47% vs. 51%), grade 3 or worse left ventricular dysfunction (22% vs. 42%), and a prescription of angiotensin-converting enzyme inhibitors (42% vs. 63%) compared with their NCH/NSA counterparts. In contrast, South Asian patients more frequently had a past history of an MI (70% vs. 52%), 3 occluded/stenosed coronary vessels on angiogram (68% vs. 51%), and treatment with coronary revascularizations (55% vs. 40%) compared with NCH/NSA patients.
Conclusions Our study demonstrates important differences in comorbid conditions, clinical characteristics, and treatment patterns among Chinese and South Asian patients compared with NCH/NSA patients with heart failure. Awareness of these differences will help to develop differential strategies necessary to prevent and manage heart failure among ethnic minority groups.
Heart failure is a major global health concern as it leads to significant morbidity and mortality. Over 23 million people worldwide live with heart failure (1), and its prevalence continues to rise, in part due to the aging population and advances in treatment of cardiovascular diseases leading to better survival (2–4). Historically, investigations in the field of cardiovascular diseases have predominantly been focused on white populations. With the rapid rise in the numbers of ethnic minorities and enhanced diversity in western countries, including the United States, United Kingdom, and Canada, awareness of the differences in cardiovascular diseases and its risk factors between ethnic groups has become increasingly important.
There is accumulating evidence to suggest that different ethnic groups may have differential risk factors and etiologies for their heart failure (5–10). Studies from the United States have shown that African-Americans have a higher incidence of heart failure than individuals of other ethnic groups, and the etiology of heart failure is less likely to be coronary artery disease (CAD) compared with their white counterparts (11,12). South Asians (SAs) were found to be at increased risk of developing heart failure due to premature CAD (13), whereas the leading cause of heart failure in the Chinese population was thought to be hypertension (14,15). Furthermore, studies in the United Kingdom showed that hospital admission rates for heart failure and presence of diabetes were higher in SA patients than white patients (16). Although prior studies have suggested different risk profiles among ethnic groups, little is known regarding the clinical profiles of ethnic minority groups with a confirmed diagnosis of heart failure. These published studies have attempted to identify differential risk profiles among ethnic groups; however, the results must be interpreted with caution due to the limitations of relying solely on administrative data. Despite these limitations, epidemiological studies on heart failure are a key to driving future prospective studies.
A large proportion of the population in Canada is comprised of visible minorities, which has surpassed 5 million people, or 16% of the overall population (17). Although prior studies have suggested different coronary risk profiles among ethnic groups, little is known regarding the clinical profiles of ethnic minority groups with a confirmed diagnosis of heart failure. With SA and Chinese populations being the first and second largest visible minority groups in Canada, respectively (17), the objective of this present study was, therefore, to compare the clinical characteristics of these 2 ethnic groups with non-Chinese/non-South Asian (NCH/NSA) patients managed in 2 specialized heart failure clinics in Ontario, Canada.
Detailed patient medical records from January 1, 2000, to January 31, 2011, were obtained and reviewed from 2 specialized heart failure clinics in Ontario, Canada (St. Michael’s Hospital and William Osler Health System) for this cross-sectional study. Both clinics follow and manage large numbers of Chinese and SA patients with heart failure.
Our cohort of 1,671 patients was comprised of Ontario residents age 19 years or older that were treated at either of the 2 clinics with a confirmed diagnosis of heart failure. In the event of missing ethnicity data in the medical records, 2 previously-validated lists of SA (including individuals from India, Sri Lanka, Pakistan, and Bangladesh) and Chinese surnames were used to identify the ethnicity of the patients (18,19). Using surname lists, Shah et al. (19) demonstrated a positive predictive value of 89.3% and 91.9% for SA and Chinese surnames, respectively. All other individuals, the great majority of whom were white patients, were grouped as the comparison group comprised of NCH/NSA patients. The patients were categorized into 3 groups: Chinese, SA, and NCH/NSA.
Patient age and sex were obtained as demographic data. In the medical history of each patient, we identified the presence or absence of CAD, angina symptoms, myocardial infarction (MI), valvulopathy, valve surgery, and coronary artery bypass graft (CABG) surgery/percutaneous coronary intervention (PCI). Prevalence of risk factors and relevant comorbid conditions, including hypertension, diabetes mellitus, dyslipidemia, peripheral vascular disease, chronic obstructive pulmonary disease, cancer, stroke, kidney disease, and ethanol consumption, were also determined. Finally, use of medical devices, including implantable cardioverter-defibrillators (ICD) and pacemakers, were documented.
Using echocardiographic data, we analyzed left ventricular ejection fraction (LVEF) and right ventricular systolic pressure. Furthermore, we categorized the patients’ left ventricular (LV) function into LVEF ≥45% and LVEF <45%, and for the purposes of this study, we only examined the differences in the incidence of Grade 3 LV dysfunction (30% to 39% LVEF) or worse severity in our patients. Electrocardiographic (ECG) data were collected to assess for the presence of atrial fibrillation and QRS durations. Angiographic data to examine the number of occluded/stenosed coronary arteries were documented, when available. We also examined laboratory data, including estimated glomerular filtration rate, urea, creatinine, high-density lipoprotein, total cholesterol, and N-terminal pro–B-type natriuretic peptide (NT-proBNP) levels (pg/ml), when available.
Finally, the medications that the patients were prescribed for treatment of their heart failure were recorded; these included diuretics, digitalis, angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blocker (ARB), beta-blockers, and aldosterone antagonists. In addition, the use of other medications such as acetylsalicylic acid, statins, calcium-channel blockers, and warfarin were examined.
We used the statistical software SAS 2013 for Windows (SAS Institute Inc., Cary, North Carolina). A p value ≤0.05 was accepted as statistically significant. Data were reported as the percentage of patients, where indicated. Rates were analyzed by the chi-square test for comparison between Chinese versus NCH/NSA as well as SA versus NCH/NSA. Continuous data were analyzed using the independent samples t test.
We used logistic regression to examine the association of the exposed variables: history of MI, diabetes, and stroke, which were 3 of the most significant medical conditions identified in this study. These conditions influenced the choice of agents and outcome variable, which is the use of ACE inhibitors and/or ARBs. A full model was constructed, including the exposure, outcome, and ethnic background. From our final model, we obtained odds ratios and 95% confidence intervals as well as p values.
Detailed medical records from January 1, 2000, to January 31, 2011, of 1,671 patients being managed by the 2 specialized heart failure clinics in Ontario, St. Michael’s Hospital and William Osler Health System, were assessed. We identified these patients as either Chinese (n = 181, 11%), SA (n = 215, 13%), or NCH/NSA (n = 1,275, 76%). The demographic data of the 3 groups are shown in Table 1. Among the patients reviewed, the mean age and sex distribution was comparable between the groups. Interestingly, there was a trend toward Chinese patients with heart failure being treated at an older age and SA patients being treated at a younger age compared with NCH/NSA patients.
Medical history, risk factors, and comorbid conditions
Data on medical history are also shown in Table 1. When compared with NCH/NSA patients, fewer Chinese patients had a history of MI (30% vs. 52%, p = 0.0014), whereas more SA patients had a previous MI (70% vs. 52%, p = 0.0025). Indeed, we observed that Chinese patients reported a history of CAD less frequently (50% vs. 69%), whereas more SA patients were found to have a history of angina (25% vs. 17%, p = 0.0144) compared with their NCH/NSA counterparts.
With regards to previous surgeries, there were significantly fewer Chinese (7% vs. 13%, p = 0.0143) and SA (8% vs. 13%, p = 0.0059) patients that had undergone valve surgery compared with NCH/NSA patients. Furthermore, compared with the NCH/NSA cohort, CABG surgery and/or PCI was done less frequently in Chinese patients (24% vs. 40%, p = 0.0014), whereas it was performed more often in SA patients (55% vs. 40%, p = 0.0021). A much smaller percentage of Chinese patients had ICDs (2% vs. 7%, p = 0.0005) as well as pacemakers (7% vs. 13%, p = 0.0024) when compared with NCH/NSA patients. Similarly, fewer SA patients had pacemakers compared with the NCH/NSA cohort (8% vs. 14%, p = 0.0176). Collectively, in the Chinese and SA groups, a combination of ICDs and pacemakers was also less often used compared with the NCH/NSA group.
When analyzed for differences in the prevalence of risk factors and comorbid conditions, there were significantly fewer Chinese and SA patients with peripheral vascular disease, chronic obstructive pulmonary disease, and cancer compared with NCH/NSA patients. In addition, ethanol consumption (1% vs. 11%, p = 0.0001) and dyslipidemia (40% vs. 52%, p = 0.0169) were less frequent in Chinese patients compared with NCH/NSA patients. SA patients also had fewer strokes (10% vs. 17%, p = 0.0444) compared with the NCH/NSA group. There were no differences found in the prevalence of hypertension, diabetes mellitus, and kidney disease between the groups.
Echocardiogram, electrocardiogram, and laboratory data
As shown in Table 2, echocardiographic data of LVEF and right ventricular systolic pressure data were also collected. We observed that Chinese patients had better average LVEF compared with the NCH/NSA group (37.5% vs. 33.7% LVEF, p = 0.04). However, we did not detect a statistically significant difference in the number of Chinese patients with preserved LV function (LVEF ≥45%) compared with NCH/NSA patients, although Chinese patients appeared to have a trend toward preserved LV function. However, our results showed that there were fewer Chinese patients that had grade III LV dysfunction or worse (22% vs. 42%, p < 0.0001) compared with NCH/NSA patients.
When the ECG data were examined, fewer Chinese (27% vs. 45%, p < 0.0001) and SA (28% vs. 45%, p < 0.0001) patients had a QRS duration >120 ms compared with the NCH/NSA patients, indicating that preserved intraventricular conduction is more prevalent in Chinese and SA patients. In addition, we found that a smaller percentage of SA patients had atrial fibrillation in their ECG compared with the NCH/NSA group (25% vs. 45%, p < 0.005).
Coronary angiogram analyses showed that fewer Chinese patients (47% vs. 51%, p < 0.0001) and more SA patients (68% vs. 51%, p < 0.01) had 3 occluded/stenosed arteries compared with their NCH/NSA counterparts.
Laboratory results (Table 3) showed that median NT-proBNP levels were significantly lower in the Chinese cohort compared with the NCH/NSA group (1,043 pg/ml vs. 3,636 pg/ml, p < 0.0001). It is important to note that only 18% of SA patients in this study had recorded NT-proBNP levels. Furthermore, Chinese patients were found to have higher mean high-density lipoprotein and urea levels, whereas SA patients were found to have lower urea and creatinine levels compared with NCH/NSA patients.
In terms of medications, we found that fewer Chinese patients were prescribed ACE inhibitors compared with the NCH/NSA group (48% vs. 67%, p < 0.001) (Table 4). On the other hand, prescriptions of ARB were more prevalent in both Chinese (35% vs. 15%, p < 0.0001) and SA patients (27% vs. 15%, p < 0.001) than in their NCH/NSA counterparts. In terms of other medications, Chinese patients were less frequently prescribed furosemide and spironolactone, more frequently on statins (67% vs. 59%, p = 0.0428), and calcium-channel blockers (33% vs. 23%, p = 0.0030) compared with NCH/NSA patients. A greater proportion of SA patients were prescribed beta-blockers (84% vs. 78%, p = 0.0208), acetylsalicylic acid (65% vs. 51%, p = 0.0002), and statins (72% vs. 59%, p = 0.0007), whereas fewer SA patients were prescribed warfarin (24% vs. 43%, p < 0.001) compared with the NCH/NSA group. The frequency of digoxin prescriptions was similar between the groups.
Logistic regression analyses showed that Chinese patients with MI are 4.37 times more likely to use ACE inhibitors than NCH/NSA patients with MI (p = 0.04), and SA patients with MI are 41 times more likely to use ACE inhibitors than NCH/NSA patients with MI (p = 0.009). With regard to diabetes, Chinese patients are 3.9 times more likely to use ACE inhibitors than NCH/NSA patients (p = 0.009), whereas SA patients are 46 times more likely to use ACE inhibitors than NCH/NSA patients (p = 0.007). However, there was no significant association with a history of stroke between Chinese and NCH/NSA patients (p = 0.3478) or between SA and NCH/NSA patients (p = 0.1569) in terms of ACE inhibitor use.
For ARB use, we have found that Chinese patients with MI are 2.7 times more likely to use ARBs than NCH/NSA patients with MI (p = 0.005). Chinese patients with diabetes are 3.2 times more likely to use ARBs than NCH/NSA patients with diabetes (p = 0.0001). In terms of a history of stroke, Chinese and SA patients are 6.8 and 4.0 times more likely to use ARBs than NCH/NSA patients, respectively (p = 0.0009 and p = 0.0167, respectively).
This present analysis was prompted in part by the continuing rise in the proportion of ethnic minority groups in Western countries as well as the accumulating evidence suggesting that there are ethnic differences in the risk for cardiovascular disease (5–10). We studied the 2 largest populations of ethnic minorities in Canada: individuals of SA and Chinese descent. Whereas previous studies that addressed ethnic differences in patients with heart failure have involved data extrapolated from administrative databases (10,16,20–22), to our knowledge, our study is the first to compare the clinical characteristics of both Chinese and SA groups against NCH/NSA patients with documented heart failure.
Previous reports have suggested that there are significant ethnic variations in the utilization of healthcare services due to either access or socioeconomic status (23,24), which may themselves be independent determinants of health. Therefore, in our study, we examined data from administrative records of heart failure patients who were managed by specialists in a hospital outpatient setting, located within similar social macroenvironments and a common universal access healthcare system.
The Chinese population
The findings of this study showed that, compared with NCH/NSA patients, there was a lower frequency of a history of MI and CAD in Chinese patients. In keeping with these results, angiographic analyses showed that there were significantly fewer Chinese patients with 3 occluded/stenosed coronary vessels compared with NCH/NSA patients. The lower rate of MI in Chinese patients found in our study is similar to the results in a previously-published international study, which reported a substantially lower incidence of acute MI in Chinese populations globally (25). Our analyses also showed that fewer Chinese patients had undergone CABG/PCI compared with the NCH/NSA group, which corresponds to the aforementioned lower incidence of MI in this population.
Hypertension was found to be the most important identifiable risk factor for Chinese patients (14), which can lead to heart failure with preserved systolic function (26). In fact, fewer Chinese patients had grade 3 LV dysfunction or worse in our study population, and there was a trend toward more Chinese patients with LVEF ≥45% compared with the NCH/NSA group. Previous reports from Hong Kong showed that 66% of Chinese patients with a clinical diagnosis of heart failure had preserved LVEF (27). In addition, a published study from a tertiary care center in Toronto, Canada, reported that a greater proportion of Chinese patients have heart failure with preserved LVEF than white patients (28). In our study, hypertension was documented as a comorbid condition, not as an etiological factor for the heart failure. We did not observe a statistically significant difference in hypertension between Chinese and NCH/NSA patients. This may be due to the fact that previous studies have compared Chinese to white populations (15), whereas our comparison group was a mix of different ethnic groups, excluding Chinese and SA patients.
In the present study, we reported that fewer Chinese patients were prescribed ACE inhibitors and more were prescribed ARBs compared with NCH/NSA patients. There have been previous studies showing a high incidence of reported cough in Chinese patients on ACE inhibitors (29–31). However, many of these studies neither had a large enough sample size nor compared Chinese with non-Chinese patients concurrently. There are 2 possible reasons for the lower incidence of ACE inhibitor prescriptions for Chinese patients: 1) healthcare providers are denying the initiation of ACE inhibitors based on anticipated intolerance; or 2) low adherence to the medications by Chinese patients. Given the unequivocal data demonstrating the benefit of ACE inhibition in heart failure patients, Chinese patients should be initially prescribed ACE inhibitors, and only substituted if intolerance develops (32).
The SA populations
In contrast to Chinese patients, a greater proportion of SA patients with previous MI and a history of angina were observed compared with NCH/NSA patients. This is also in line with previous international studies as well as studies in other Western countries, including the United Kingdom (24). Furthermore, our angiogram data analyses showed that there were significantly more SA patients with 3 occluded/stenosed coronary arteries compared with the NCH/NSA cohort, which is in agreement with our data demonstrating that more SA patients have been treated with CABG/PCI.
A retrospective sequential chart review of SA and white individuals with a diagnosis of HF at 2 Toronto-area hospitals showed that SA patients were younger and more often diabetic (20). The results of our study did not show a significant difference in diabetes between the SA and NCH/NSA populations; the discrepancy in the findings is likely due to our reliance on administrative data, which may not have all relevant comorbid illnesses documented. However, SA patients did not experience increased mortality (20). The reasons for these inverse relationships in mortality in these populations may be attributable to the incidence of atrial fibrillation post-MI. Post-MI prognosis has been shown to be worse among patients complicated by atrial fibrillation (33). Our results showed that SA patients with heart failure were found to have the lowest prevalence of atrial fibrillation on ECG and not on their history, which is a comorbidity/risk factor for heart failure. Previous studies conducted in Alberta as well as the Toronto area in Canada showed similar results of lower rates of atrial fibrillation at presentation in SA patients (22,23). The explanation for the lower rates of atrial fibrillation is not fully understood and warrants further investigation.
Our study results are complimentary to those derived from earlier administrative studies. The limitations of administrative data deserve consideration. First of all, the system that was used to group patients based on ethnicity rather than using self-reported ethnicity may have led to misclassification of certain patients. Second, the discrepancy in some of our findings lacking statistical significance is likely due to our reliance on administrative data, which may not have all relevant parameters documented. Third, as with all retrospective data collection, ECG and laboratory data were missing for some patients, although it appeared to be proportionate between the groups. Fourth, because Toronto is the largest and most diverse city in Canada, the results may not apply to the rural areas of Canada in terms of population demographics, clinical characteristics, and treatment of heart failure patients. Despite these limitations, our study is the first to compare the clinical characteristics of both Chinese and SA patients against NCH/NSA patients with a confirmed diagnosis of heart failure managed within similar social macroenvironments and a common healthcare system.
There are currently no large-scale randomized controlled trials of managing heart failure specifically in the Chinese and SA populations. Much of the published guidelines for therapy are results of studies on white populations. Thus, the need for the understanding of ethnic variations in heart failure is necessary for better management of ethnic minority patients, keeping in mind the ethnic differences in cultural beliefs, health literacy, and risk tolerance. There are significant differences in the clinical characteristics of heart failure among Chinese, SA, and NCH/NSA patients. The results of this study are valuable in fostering the awareness of the unique risk profiles among the largest and increasing ethnic minority groups in Canada so that appropriate screening procedures and support programs can be implemented.
This study was supported by the Mandarin Buffet through a designated grant program of the Heart and Stroke Foundation of Canadahttp://dx.doi.org/10.13039/501100000222. All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- angiotensin-converting enzyme
- angiotensin II receptor blocker
- coronary artery bypass graft
- coronary artery disease
- implantable cardioverter-defibrillator
- left ventricular ejection fraction
- myocardial infarction
- non-Chinese/non-South Asian
- N-terminal pro–B-type natriuretic peptide
- percutaneous coronary intervention
- South Asian
- Received December 2, 2013.
- Revision received February 4, 2014.
- Accepted February 8, 2014.
- American College of Cardiology Foundation
- McMurray J.J.,
- Petrie M.C.,
- Murdoch D.R.,
- Davie A.P.
- Ezekowitz J.A.,
- Kaul P.,
- Bakal J.,
- Armstrong P.W.,
- Welsh R.C.,
- McAlister R.A.
- Velagaleti R.S.,
- Pencina M.J.,
- Murabito J.M.,
- et al.
- Liang C.S.,
- Delehanty J.D.
- Newton J.D.,
- Blackledge H.M.,
- Squire I.B.
- Sanderson J.E.,
- Tse T.F.
- Blackledge H.M.,
- Newton J.,
- Squire I.B.
- Statistics Canada
- Choi B.C.,
- Hanley A.J.,
- Holowaty E.J.,
- Dale D.
- Kaul P.,
- McAllister F.A.,
- Ezekowitz J.A.,
- Grover V.K.,
- Quan H.
- Khan N.A.,
- Grubisic M.,
- Hemmelgarn B.,
- Humphries K.,
- King K.M.,
- Quan H.
- Smedley B.D.,
- Stith A.Y.,
- Nelson A.R.
- Quan H.,
- Fong A.,
- De Coaster C.,
- et al.
- Tunstall-Pedoe H.,
- Kuulasmaa K.,
- Mähönen M.,
- Tolonen H.,
- Ruokokoski E.,
- Amouyel P.
- Vasan R.S.,
- Larson M.G.,
- Benjamin E.J.,
- Evans J.C.,
- Reiss C.K.,
- Levy D.
- Woo J.,
- Chan T.Y.
- Howlett J.G.,
- McKelvie R.S.,
- Costigan J.,
- et al.
- Crenshaw B.S.,
- Ward S.R.,
- Granger C.B.,
- Stebbins A.L.,
- Topol E.J.,
- Califf R.M.