Article Text

other Versions

Hypertension and cardiovascular disease endpoints by ethnic group: the promise of data linkage
  1. Charles Agyemang1,
  2. Raj Bhopal2
  1. 1Department of Public Health, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands
  2. 2Edinburgh Ethnicity and Health Research Group, Centre for Population Health Sciences, University of Edinburgh, Edinburgh, UK
  1. Correspondence to Dr Charles Agyemang, Department of Public Health, Academic Medical Centre, University of Amsterdam, Amsterdam 1105 AZ, The Netherlands; c.o.agyemang{at}

Statistics from

Hypertension is the most important risk factor for cardiovascular diseases (CVD), accounting for approximately 45% of global CVD morbidity and mortality.1 Evidence suggests striking differences in blood pressure (BP) and hypertension prevalence between ethnic groups. West African descent adults living in Europe and North America, whether they come directly from Africa or indirectly from the Caribbean, generally have higher BP levels and a higher prevalence of hypertension than European descent populations (henceforth, white individuals), with this being seen at all ages in North America and only from adulthood in the UK.2 ,3 Chinese-origin people also have slightly higher BP and prevalence of hypertension than white individuals.4 ,5 The evidence is mixed when it comes to the South-Asian descent populations (ie, Indian, Pakistani, Bangladeshi and Sri Lankan people). In a systematic review in the UK, BP levels among South-Asian individuals were generally similar to that of the UK general population, but there were stark differences among the South-Asian subgroups, with slightly higher BP in Indian individuals, slightly lower BP in Pakistani individuals, and much lower BP in Bangladeshi individuals.6 Studies in The Netherlands7 and Canada,5 ,8 however, show a higher hypertension prevalence in South-Asian than in white individuals. In the Ontario Health Survey, the age-standardised hypertension prevalence among South-Asian individuals was 30.1% compared with 20.7% among white Canadian people.8 South-Asian were still more likely than white Canadian individuals to have hypertension even after adjustment for age, sex and body mass index.

While hypertension remains the most important risk factor for CVD, its contribution to the ethnic differences in CVD outcomes is still sometimes puzzling. In the UK, although the BP levels are similar or lower in the South-Asian relative to the general population, they have a higher mortality from stroke and ischaemic heart disease. There is an excess of 50–70% in ischaemic heart disease mortality rates in South-Asian individuals in the UK with the highest rates in Bangladeshi people,9 and yet Bangladeshi individuals have the lowest mean BP and hypertension prevalence among the South-Asian subgroups.6 This has led to a number of hypotheses10 and new empirical research on the role of hyperglycaemia.11 Similar puzzling findings are also observed in West African compared to white individuals. For instance, it is a paradox that West African descent people in the UK, despite having a high prevalence of hypertension and diabetes, have low mortality rates from coronary heart disease.9 Moreover, Chinese people have more favourable CVD, especially coronary heart disease, outcomes than white individuals, although BP levels are similar to or higher in Chinese than in white individuals.4 ,5 These kind of variations are probably due to differentials in other risk factor; for example, smoking, cholesterol and obesity.4 ,5

Older studies among South-Asian populations, for example, the Brent and Southall cross-sectional studies, indicated that other conventional cardiovascular risk factors, such as smoking, diabetes or high cholesterol in addition to hypertension could not explain the ethnic differences in CVD outcomes. Part of the inconsistencies in the relationship between risk factors and CVD outcomes among these populations may relate to the quality of published data. A review by Moe and Tu12 concluded that most published studies are based on small samples and that additional studies across different ethnic groups are needed. In addition, there is a general lack of cohort studies to provide accurate incidence and/or mortality data by ethnic group. To establish whether the CVD burden does indeed differ between ethnic groups, we need accurate population-based data on the incidence and survival of CVD by ethnic groups. Forouhi and colleagues13 have recently followed up the Brent and Southall cross-sectional sample cohorts, and this new analysis confirms that the excess of CVD in South-Asian individuals cannot be fully explained on the basis of known risk factors. Pending the availability of large-scale cohort data in South-Asian populations, such as the West London Cohort (also known as LOLIPOP) study, we are reliant on either follow-up of cross-sectional studies or retrospective cohort studies created using data linkage.

The use of linkage techniques is increasing and becoming a potentially useful source to create cohort studies to provide incidence and/or mortality data by ethnic group.14 The linkage technique poses many challenges. First, producing a retrospective cohort using the linkage technique is a difficult task, although much easier and cheaper than prospective cohort studies. Second, the approach is still in its infancy, at least among ethnic minority groups, and the precision of the findings is hard to guarantee as it depends on the quality of databases being linked. Third, there are legal hurdles in linking different databases. Finally, and crucially from a cardiovascular perspective, linkable data on risk factors are usually lacking. The strength of the linkage technique is its ability to create large sample sizes and long-term follow-up.

In this context we welcome, the study by Quan and colleagues,15 which compared ethnic and sex differences in the incidence of newly diagnosed hypertension and the subsequent risk of CVD outcomes including myocardial infarction, heart failure, stroke and mortality among South-Asian, Chinese and white patients. The authors used four routinely collected administrative datasets including hospital discharge abstracts, physician claims, population registry and vital statistics registries from 1994 to 2005 from the Canadian provinces of British Columbia and Alberta. These databases were linked using unique personal identifiers or personal information. The authors found that Chinese patients had both a lower incidence of hypertension and a lower risk of developing cardiovascular endpoints compared to white patients. This was true for both men and women when the analyses were stratified by sex and were adjusted for important available covariates. By contrast, South-Asian patients had a higher incidence of hypertension than white Canadian individuals. Despite the higher hypertension incidence, South-Asian patients had lower CVD endpoints compared to white Canadian patients. The differences hardly changed after adjustments for important factors. The differences in CVD endpoints largely remained after sex stratification and adjustment for potential confounders except for the incidence of heart failure, which was similar between South-Asian women and white Canadian women.

Like previous linkage analyses, this study provides a practical strategy and new data to shed light on the increasingly important issue of ethnic differences in CVD incidence. The findings of the lower incidence of CVD outcomes among Chinese Canadian compared to white Canadian individuals are consistent with earlier studies in Canada and the UK.5 ,9 The findings of the higher incidence of hypertension among South-Asian people in British Columbia and Alberta is also consistent with the higher prevalence rate found among South-Asian individuals in Ontario, but the lower rates of CVD endpoints found among South-Asian people in this current study do not corroborate with the previous work.5 ,9 Earlier studies show higher rates of CVD outcomes in South-Asian than in white populations.5 ,9 Most risk factors for CVD have also been shown to be higher in South-Asian than in white individuals in Canada.5 ,8 In the SHARE study, South-Asian people had an increased prevalence of glucose intolerance, higher total and low-density lipoprotein cholesterol, higher triglycerides and lower high-density lipoprotein cholesterol, and much greater abnormalities in novel risk factors including higher concentrations of fibrinogen, homocysteine, lipoprotein (a), and plasminogen activator inhibitor 1.5

Given the current knowledge on CVD and risk factors in Canada and elsewhere in Europe and North America, might it be that the current finding of the lower CVD endpoints in South-Asian individuals is a statistical artefact? Corroboration is required before we can judge on this. First of all, the favourable CVD endpoints found in Chinese people and the high hypertension incidence in South-Asian individuals are in line with other studies in Canada. In addition, several potential explanations might underlie the current findings including the ‘migrant pond effect’. The migrant pond effect hypothesis postulates that the health status of the local reference population will ultimately determine the relative health standing of the ethnic minority and migrant groups within the country or locality in which they live.16 The current studies were undertaken in British Columbia and Alberta, which might be different from other provinces in Canada where earlier studies were carried out. Tanuseputro and colleagues17 found large variations in the prevalence of CVD risk factors and smoking-attributable mortality estimates across geographical regions in Canada. A recent study by Bhopal and colleagues,18 in which they compared circulatory disease mortality in the same country of birth group across European countries, found substantial between-country differences. For example, Chinese-born people living in France had lower circulatory mortality rates than Chinese-born counterparts living in different parts of Europe. One study in Australia also found a lower circulatory disease mortality in South-Asian than in Australian-born people,19 contrary to previous studies in Europe and North America.5 ,9 Furthermore, a recent nationwide linkage analyses in Scotland also found important ethnic differences in heart failure,14 but the differences did not confirm previous work indicating several-fold differences.13 So, context is important, and might lead to genuine differences in the risk factor–disease outcome relationships in different ethnic groups in different places.

The future for linkage-based studies among ethnic minority groups looks promising, but there is a need for internationally coordinated validation studies to establish the reliability of the linkage processes and quality and comparability of databases.

Over the next few years new results from the growing number of linkage studies, together with new follow-ups of cross-sectional surveys, and above all prospective data from recently established cohorts, will lead to a critical re-evaluation of previous conclusions on ethnic variation in CVD outcomes.


View Abstract


  • Competing interests None.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles