Other Studies

• Identifying individuals at high risk of CVD is one of the main objectives of primary prevention and the first step towards treating high-risk patients with modifiable risk factors.
• When evaluating a patient’s global CHD risk, it is important to consider all risk factors simultaneously.
• The SCORE risk estimation system directly estimates total fatal CVD risk for clinical management of CVD risk in European populations.
• The UKPDS risk engine is a good CHD risk estimation model for primary prevention of CHD in type 2 diabetic patients.
• CUORE has produced a predictive equation specific to the Italian population.

Identifying subjects at high risk of cardiovascular disease (CVD) is a key priority of primary prevention and the first step towards reducing CVD risk by targeting modifiable risk factors [1]. Traditionally, risk factor guidelines have focused on single factor assessment rather than on individual’s global risk based on a combination of risk factors [2]. However, large prospective studies [3,4] have clearly shown that coronary heart disease (CHD) is a multifactorial disease and that coronary risk factors do not act in isolation but rather in conjunction with each other. It is therefore important to consider all risk factors when assessing a patient’s global CHD risk. This way of identifying high-risk individuals recognizes not only CHD’s multifaceted nature and the multiplicative effect of risk factors but also the importance of assessing and managing patients beyond mere treatment of risk factors [2].

Over the last decades, numerous epidemiological studies [3,4] have focused their attention on global absolute risk for identifying high-risk individuals. Several risk prediction methods have been developed to identify individuals at high risk of CHD or CVD. Among those risk prediction methods, the Framingham Heart Study [3] and the PROCAM study [4] are two well-known examples of predictive equations derived from large prospective studies. Additional risk charts and scores have also been developed from other widely known studies such as the Systematic COronary Risk Evaluation (SCORE) project [5], the United Kingdom Prospective Diabetes Study (UKPDS) [6], and the Italian CUORE project [7]. All these risk charts and scores are useful tools for evaluating an individual’s global CHD risk in clinical practice.

To date, one of the best-known risk prediction systems is based on the Framingham Heart Study [3]. However, the Framingham Study examined a relatively homogenous American population living within a limited geographical area (the city of Framingham, Massachusetts). In view of these limitations, the European Society of Cardiology and the Second Joint Task Force developed a risk estimation system based on a large pool of European studies to capture regional variation in CHD risk [5]. Their efforts led to the SCORE project, which assembled a pooled dataset of cohort studies [8-22] from 12 European countries. The consolidated database included 205,178 individuals (117,098 men and 80,080 women) without previous history of heart attack. Among the subjects, there were 7,934 cardiovascular deaths, of which 5,652 were attributed to CHD. Ten-year risk of fatal CVD was calculated using a Weibull model. The risk of cardiovascular death was calculated by combining two separate estimation equations: a model for CHD and a model for non-coronary atherosclerotic CVD.

The endpoint for the SCORE project was fatal atherosclerotic CVD. In interpreting CVD risk estimates, the SCORE project used fatal CVD only, instead of a combination of fatal and non-fatal events. This was because of limited availability of non-fatal endpoint data in several cohort studies and possible variation in endpoint definition. The SCORE project also provides specific European risk charts for countries at high and low cardiovascular risk, using cardiovascular mortality as the endpoint. The baseline survival functions for the cohorts from Denmark, Finland, and Norway were used to develop the high-risk model while baseline survival functions of studies from Belgium, Italy, and Spain were used to develop the low-risk region model. Risk was calculated for two different risk charts. The first chart was based on total cholesterol and the second on the cholesterol/HDL cholesterol ratio. In each chart, the risk factors used were sex, systolic blood pressure, and smoking habits. Age was used as a measure of exposure time to risk rather than as a risk factor. Compared to previous models from the European Society of Cardiology and the Second Joint Task Force [2], SCORE risk charts did not include risk for age 30 because individuals 30 years of age are essentially risk-free for the next 10 years and because there were no events in this age group. The SCORE risk charts therefore provide information for the 35 to 65 age group, which is when risk changes most rapidly.

Compared to the recommendations of the Second Joint Task Force of European Societies [2], which propose separate risk charts for type 2 diabetics, the SCORE project includes diabetic subjects, for whom no separate risk charts have been developed. Moreover, unlike Framingham and PROCAM algorithms that consider type 2 diabetes in score calculation, SCORE risk functions do not include a dichotomous diabetes variable even though diabetes is known to markedly increase CVD risk [2]. The World Health Organization Multinational Study Group on Vascular Disease in Diabetes [23] has stated that the assessment of CVD risk in diabetes must include “diabetes-related” variables as well as conventional risk factors. Risk estimations should also be derived from large representative groups of diabetic subjects with homogeneous baseline data. Accordingly, UKPDS [6] has developed models for estimating absolute CHD risk in newly diagnosed type 2 diabetic men and women. When evaluating CVD risk in diabetic patients using the SCORE project, the CVD risk at every risk factor combination will be at least twice as high in diabetic men and up to four times higher in diabetic women when compared to the risk given in the SCORE charts.

The SCORE project therefore provides an “average” European chart for countries with no cohort data. It aims to promote the production of national cardiovascular risk charts using cardiovascular mortality data and SCORE risk functions with appropriate adjustments.

For more information, please visit the European Society of Cardiology website.

As previously stated, individuals with type 2 diabetes are at increased CHD risk compared to the general population [24-26]. It has been suggested that diabetic individuals with no CHD history have a similar risk of developing an acute myocardial infarction (MI) as nondiabetic individuals who have had a previous MI [27]. Unfortunately, CHD prediction models such as Framingham risk equations [3] were not specifically designed for type 2 diabetic patients. These algorithms tend to underestimate CHD risk in individuals with type 2 diabetes [28]. Moreover, these models do not consider diabetes-specific risk factors such as glycemic control and duration of diabetes. Accordingly, the UKPDS [6,29] developed an equation for estimating the risk of new CHD events in men and women with type 2 diabetes. Unlike previous risk equations, the UKPDS risk model is diabetes-specific and incorporates glycemia.

The UKPDS cohort was composed of 4,540 patients with newly diagnosed type 2 diabetes. The inclusion criteria also encompassed fasting plasma glucose greater than 6 mmol/l and no previous history of MI, angina, or heart failure. The patients were recruited between 1977 and 1991 and underwent a follow-up of 10.7 years (median). In the model, CHD was defined as the incidence of fatal or non-fatal MI or sudden death. Baseline risk factors were age at diagnosis of diabetes, sex, ethnic group (Afro-Caribbean, Caucasian, or Asian-Indian), smoking status, HBA1c, systolic blood pressure, total cholesterol/HDL cholesterol ratio, and time since diagnosis of diabetes. For some variables, such as HBA1c, systolic blood pressure, and cholesterol/HDL cholesterol levels, year 1 and 2 measurements were averaged to improve model stability.

For CHD estimation, a fully parametric model combining hazards ratio and absolute events rates in a single equation was used to estimate events rates and survival probabilities. Health planners can use the UKPDS model to estimate resource use, perform power calculations in clinical trials, and estimate effectiveness and cost-effectiveness early in drug development cycles.

In comparison with previous models that do not include any measure of glycemia, an important feature of the UKPDS model is the inclusion of HBA1c as a continuous CHD risk factor. A direct causal relationship between glycemia measured by HBA1c, and CHD has not yet been proven [30], but it is important to acknowledge that a predictive relationship does exist [31,32]. The UKPDS prediction model also includes some important CHD risk factors such as blood pressure [33] and total cholesterol/HDL cholesterol ratio [34]. The UKPDS risk engine is therefore a good model for estimating CHD risk for the primary prevention of CHD in type 2 diabetic patients.

For more information, please visit the UK Prospective Diabetes Study website.

In Italy, CVD is the leading cause of death in adults and accounts for 44% of total deaths [35,36]. Risk prediction models are therefore crucial to estimate CVD risk and identify individuals at high risk of CVD. The CUORE study is a large prospective cohort follow-up study, including cohorts from the Northwest, Northeast, Centre, and South of Italy. The aim of the CUORE project [7] was to develop a 10 year coronary risk predictive equation specific to the Italian population. Among the sample of 6,865 men 35 to 69 years of age and CHD-free at baseline, 312 first fatal and non-fatal major coronary events occurred in the 9.1 year median follow-up. Women were excluded from the analysis because of the small number of events and the shorter follow-up period. The follow-up period is currently being extended and more stable estimates should later be available for women.

The CUORE predictive equation includes well-known CHD risk factors such as age, total cholesterol, systolic blood pressure, cigarette smoking, HDL cholesterol, type 2 diabetes, hypertension medications, and family history of CHD. In the CUORE equation, body mass index (BMI), an index of overall obesity, was not an independent risk factor and did not improve CHD prediction. This could be explained by the fact that other obesity-related risk factors, such as type 2 diabetes, elevated total cholesterol, high blood pressure, and low HDL cholesterol, were already considered in the model. BMI’s impact was thus minor once these risk factors were taken into account. Triglyceride levels were also not included in the model because their inclusion eliminated the protective effect of HDL cholesterol. Once HDL cholesterol was included, adding triglycerides to the model did not improve CHD prediction [37].

By not overestimating risk, the CUORE equation has been shown to be more accurate in predicting CHD events than well-known equations such as Framingham and PROCAM. The CUORE prediction equation may therefore help distinguish adult men at high CHD risk from lower-risk populations. However, this CHD risk equation requires further study in low-risk populations.

In recent years, interest has risen in global estimation of CHD and CVD risk, and efforts have been made to create prediction tools to identify high-risk individuals. Global cardiovascular risk can be calculated using several CHD risk charts and scores derived from large prospective studies in North America and Europe. Cardiovascular risk charts and individual scores should be simple and easy to use in clinical practice so that high-risk individuals can be quickly identified. These risk prediction systems are crucial for effectively managing CVD. Globally, these prediction tools all have the same objective: identify high-risk patients in order to treat their modifiable risk factors, ultimately decreasing mortality and increasing life expectancy. However, despite all these efforts, CVD still remains the primary cause of death in many developed countries, which underscores the need to develop or improve CVD risk prediction models for optimal assessment of global CVD risk.