Assessment of Cardiovascular Disease Risk in South Asian Populations

3.1. CVD Risk Profiles in South Asians Compared with Those in Caucasians and South Asians Living in Other Countries

The diversity and differences in CVD risk factors in native South Asians compared to South Asians who have migrated abroad and between the South Asians and other ethnic groups (Caucasians and Chinese) are shown in Table 1. There were marked differences in the distribution of CVD risk factors [5–7]. It was concluded that the wide array of differences in CVD risk profile makes combining data from different misleading South Asian countries [6, 7]. Even within countries such as India, marked variations in risk profiles due to difference in culture, lifestyle, and religion have been reported. Nevertheless, all of these studies have concluded that despite their lower body mass index (BMI), body weight, and serum cholesterol, South Asians have greater evidence of other features of dyslipidaemia including lipoprotein(a) and other CVD risk factors than Caucasians [5–10]. Additionally, South Asians had more prominent abnormalities in fibrinogen, plasminogen activator inhibitor 1, homocysteine, C-reactive protein, insulin sensitivity, and psychosocial factors (stress, depression) [6–9].

3.2. Assessment of CVD Risk Factors in South Asians in Their Native Countries

The widely discussed CVD risk factors in South Asians are listed in Table 2. The INTERHEART study and other recent researches in Singapore, the United Kingdom, and Canada, which examined large populations from South Asian countries [7, 8, 10, 11], reported that South Asians tended to experience both the “classical” and nontraditional CVD risk factors in a higher level and at a younger age than the other populations [7, 8, 11, 12]. For example, South Asians were more likely to have more severe abnormalities of harmful factors—including elevated apo B100/apo A-I ratio [7, 11, 13], lipoprotein(a) [8, 11, 14], depression, and stress at work or home [7, 13], and had lower rates of protective factors—physical activity [7, 13], regular fruits, and vegetable intake [7, 13]. Concurrently, potentially widespread environmental pollution, such as that from arsenic contamination and indoor air pollution, further complicates the situation. Arsenic exposure which is associated with vascular inflammation and endothelial dysfunction exerted by atherosclerotic plaque, oxidative stress, and lipid peroxidation [15] is a huge problem for an estimated 40–60 million people in the Ganges delta of South Asia [16] as they consume arsenic contaminated groundwater [16] and food from plant origin [17]. Most of the rural houses in Bangladesh [18] and 75% of households of India [19] use solid fuels for cooking which are the major sources of indoor air pollution and pose serious health risks including CVD and low birth weight [18, 19].

With 31% of all infants born with low birth weight, South Asians had the highest incidence of this problem [20] which has been consistently associated with increased risk of insulin resistance and atherosclerotic CVD later in life [21, 22]. It has been postulated that low birth weight contributes to the early development of the CVD risk factors, including blood pressure and plasma glucose concentrations, insulin resistance, abnormal fibrinogen/factor VII, and apolipoprotein levels [23] that are common among South Asians.

3.3. Currently Available and Validated Risk Algorithms

Table 3 summarizes details concerning the most widely used risk scores, the populations from which they were derived, the risk factors included, and the different CVD outcomes which were examined. A person’s estimated probability of developing a major CVD event(s) during the following 5-or 10-year period is known as absolute or short-term risk. Absolute risk calculators are developed and validated in longitudinal cohort studies which estimate the effect of multiple CVD risk factors on CVD events [24]. However, if any CVD risk factor was left untreated for a long time, it can produce high risk and cumulative damage [25]. Not only for this reason but also because age is the major determinant of short-term absolute risk a lifetime risk, scoring method has been developed and has been recommended for younger patients to offer additional risk burden.

The relationship between independent major risk factors and coronary heart disease (CHD) risk was first described over 50 years ago using data from the Framingham Heart study. Since then, over 100 different CVD risk scores have been developed, mainly based on findings on Caucasians in developed countries. Among the most commonly used risk scores, at least five have been derived from the Framingham study [26–28]. Other studies from which risk scores have been derived include those of the Munster group (PROCAM) [29], ARIC [30], QRISK (electronic medical database) [31, 32], Reynolds [33, 34], the Scottish Heart Health Extended Cohort [35], Strong Heart Study [36], USA-PRC (People’s Republic of China Collaborative Study of Cardiovascular Epidemiology) [37], the SCORE project [38], Personal heart [39], NHEFS [2], and the UKPDS [40]. None of the above risk scores have been derived from epidemiologic studies undertaken in developing countries and/or a South Asian population living in South Asia. Although the IHMRS (obtained from INTERHEART data) has been validated in South Asians, it has the limitation of being derived from a case-control study which might not be considered to be as robust in predicting future events as equations derived from prospective data [41].

The numbers of risk factors that incorporated into the risk scores have varied widely ranging from 4 to 14. All algorithms have included age, gender, blood pressure, and smoking status, and also they almost included lipid parameter(s) and diabetes status. Of note, lipid parameters were not included in the nonlaboratory model of the 2008 Framingham risk score [28], the PROCAM 2007 (stroke) risk score [29] or the NHEFS risk score [2]. The CVD outcome(s) predicted by the risk scores also differs markedly, as again shown in Table 3.

3.4. Calculation of CVD Risk with Currently Available Risk Tools

There is no specific risk score that has been derived from South Asian populations; nor to our knowledge has any attempt been made to calibrate either short-term or lifetime risk stratification algorithms in these populations.

Some of the currently available risk scores have been applied to calculate CVD risk in South Asians. For example, when the ATP III (Adult Treatment Panel) risk score (based on the Framingham study) was applied to a cohort of healthy Indian industrial workers of 40.8 ± 10.9 years of age on average (range 20–69 years), it was estimated that one in two men and three in four women had a low short-term risk for CVD [42]. Based on this aggregate of risk factor burden, these results highlight a key limitation—an overestimation of CVD risk [42]. As with almost all of the combinations of risk factors at their extreme values, the ATP III risk score predicts 10-year CVD risks below 10% for nonsmoking men <45 years of age and all women <65 years of age [43]. In another Indian cohort, three risk scores were applied and predicted different levels of CVD risk for the individuals in the same population [12]. The Framingham model predicted only 5.3% of the study subjects (males aged 37.2 ± 14 years and females aged 43.2 ± 13.7 years) at high risk; an apparent underestimation [12] has given that an adjusted Framingham score calculates 30-year risks to be 7.9% for women and 18.0% for men in the population from which it originated [44]. That Framingham algorithm has been observed to perform poorly across socioeconomic groups, and underestimated risk in socioeconomically deprived groups [45] may be an explanation for this underestimation. The Joint British Cardiac Society (BCS)/British Hypertension Society (BHS)/British Hyperlipidemia Association (BHA) risk score also appeared to underestimate the CVD risk [12]. However, the risk underestimation by the Framingham algorithm and the British Cardiac Society algorithm was disproportionate (5.3% by Framingham, 3.7% and 4.4% by the joint British Societies scores, resp.) [12].

4.1. Chronological Age and CVD Risk

Although chronological age partly influences the formation of atherosclerotic plaque, it does not reflect true cardiovascular age [47]. As a result, assigning the same CVD risks to all individuals of the same chronologic age will be problematic because of the great variation in plaque burden at that point in a lifetime [47]. Yet, age remains one of the key variables in CVD risk scores. The sensitivity of the Framingham risk score and SCORE was lower in the younger age group (40–45 years) and older age group (70–85 years) compared with that in the middle age group (55–64 years) [48] even when risk factor levels were unfavourable among the younger persons. For example, the Framingham score nearly reached 10% risk in people aged 18–29 years and calculated risk <12% in people aged 30–39 years despite the fact that the risk factor burden was remarkably high in these young individuals. In the INTERHEART study, hypertension, smoking, dyslipidaemia, and diabetes in women <60 years of age were more strongly associated with myocardial infarction than the presence of the same risk factors in women >60 years of age [13]. A concerning fact for South Asians is that CVD risk for this group should be estimated at least 10 years earlier than in Caucasians and Chinese as these people harbinger CVD risk factors at a younger age [7, 13]. Additionally, there may be variations in the extent of atherosclerosis due to not only “conventional” but also novel risk factors which can partly explain their higher rates of CVD [8].

4.2. Migration and Changes in CVD Risk Factors

Migration is associated with changes in lifestyle, psychosocial interaction, dietary habit, climate, culture, and living conditions, which ultimately may influence the CVD risk profile. However, the change in CVD risk depends on the country which migration has taken place from and the country to which people have migrated. Most studies have reported that migration has a deleterious effect on CVD risk factors. For example, migrants from the Indian subcontinent who settled in the USA, UK, or other European countries have been reported to have increased obesity, apolipoprotein B levels [7, 11, 49], fasting plasma insulin [11, 49], C-reactive protein [49], serum triglyceride [50], and decreased β cell functions [11, 49] compared to the natives of those countries [11, 49–51]. In addition, migration could be associated with significant psychosocial stress that may impact on CVD, though such evidence is sparse [52]. All these factors make it questionable whether the risk score derived from the Strong Heart study (undertaken in American Indians) or ARIC and UKPDS algorithms in which ethnicity is considered as one of the risk factors can be transferred directly for use in the South Asian populations.

4.3. Central Obesity and Insulin Resistance in South Asians

Several explanations including the role of body composition have been proposed to explain the paradox between the high prevalence of diabetes and atherosclerotic CVD despite a lower mean BMI and waist circumference among South Asians. Studies have shown that for a given BMI or waist circumference South Asian males had approximately 6% higher total body fat than Caucasian men [52]. Additionally, South Asians had larger abdominal subcutaneous adipocytes than Caucasians, a phenomenon associated with insulin resistance independent of waist circumference. Several longitudinal studies have demonstrated that individuals who are insulin resistant develop abnormal CVD risk factors long before the development of overt diabetes [41, 52]. These abnormal CVD risk factors include low adiponectin and altered adipokines levels, lipid abnormalities [41], glucose intolerance, elevated blood pressure [41, 52], chronic inflammation [52], and an increased procoagulant state [41] that ultimately lead to microvascular and macrovascular complications [41, 52, 53]. Risk assessment is a useful tool in CVD prevention as it promotes the most effective and cost-effective use of preventive strategies and highlights the importance of lifestyle modification. It may also be used as a motivational tool according to the manner in which it is presented. South Asian populations are well known to suffer an excessive and premature burden of CVD. But, estimating their CVD risk accurately is difficult as the popular risk algorithms either underestimate [12] or overestimate the CVD risk [42], and different algorithms often predict CVD risk differently in the same individual [12].

4.4. Strengths and Limitations

This review is a qualitative synthesis of published studies which used different epidemiological designs with varying sample sizes and did not adhere only to the hierarchy of evidence. However, the literature search was systematic, and the qualitative synthesis that was made was robust. Moreover, this study provides complete information on prevailing CVD risk factors and the vulnerability of the South Asians to CVD, the CVD risk factors difference between South Asians and Caucasians and others, and the risk factors included in the currently available risk scores and their prediction of CVD risk among these populations.