The guidelines that cover the screening of patients for elevated serum lipid levels, and the treatment of patients with lipid abnormalities, rest on calculations of individual patients’ risk for a future cardiovascular event. Broadly, risk factors are arbitrarily divided into three major categories: nonmodifiable, modifiable, and emerging. See Table 1, below.
Table 1. Basic Categories of Risk Factors for Future Cardiovascular Events (Open Table in a new window)
Nonmodifiable risk factors
Age, sex, family history, genetic predisposition
Modifiable risk factors
Smoking, atherogenic diet, alcohol intake, physical activity, dyslipidemias, hypertension, obesity, diabetes, metabolic syndrome
Emerging risk factors
Elevation in C-reactive protein (CRP); fibrinogen; coronary artery calcification (CAC); homocysteine; lipoprotein(a); small, dense LDL
An elevated total cholesterol or low density lipoprotein (LDL) cholesterol level and/or a reduced high density lipoprotein (HDL) cholesterol level are traditional risk factors for cardiovascular disease (CVD); when cholesterol elevations occur in combination with other risk factors, a much higher risk for CVD is predicted. Risk algorithms include lipid levels with some combination of the following other traditional risk factors:
Family history of premature coronary heart disease (CHD; ie, in a first-degree male relative < 55 y or female relative < 65 y)
The most commonly used risk algorithms developed with United States population cohorts include the following:
Framingham Risk Score (FRS; multiple adaptations)
Reynolds Risk Score (RRS)
American College of Cardiology/American Heart Association
arteriosclerotic cardiovascular disease risk estimator (AC/AHA-ASCVD)
Commonly used risk algorithms developed with European population cohorts include the following:
Systematic Coronary Risk Evaluation (SCORE)
Framingham Risk Score
The Framingham Risk Score (FRS) was developed in 1998 to assess the 10-year risk of coronary heart disease (CHD) for individuals with different combinations of risk factors. The data used was from the Framingham Heart Study, an ongoing study begun in 1948 of healthy adults, in a largely white population in Framingham, MA.
Since 1998, many adaptions and revisions of the score have been published. Notable versions include the following:
2002 adaption of the National Heart, Lung and Blood Institute (NHLBI) Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (FRS-ATP-III)
2008 10-year Framingham CVD risk score (FRS-CVD)
2006 Lifetime Framingham CVD Risk Score (Lifetime-FRS)
2009 30-year Framingham HCVD risk score (FRS-HCVD)
The 2002 adaption deleted the variables of diabetes and family history of premature coronary heart disease (CHD), added the impact of treatment for hypertension, and used only hard coronary heart disease endpoints in its calculations.
The 2008 adaption included additional cardiovascular events (ie, stroke, transient ischemic attack) that had not previously been assessed. The 10-year CVD risk score is expected to be higher than the 10-year CHD risk score.
The 2006 Lifetime-FRS estimates risk from age 50 based on four risk factors: total cholesterol, systolic blood pressure, cigarette smoking, and diabetes.
The 2014 International Atherosclerosis Society (IAS) guidelines recommends the 2006 Lifetime-FRS for countries where recalibration values can be applied to risk calculations.
The 2009 30-year FRS-HCVD estimates risk from age 45 based on a larger number of major risk factors and uses only hard disease endpoints.
Reynolds Risk Score
The Reynolds Risk Score (RRS) was developed in 2007 with data from a 10-year study of 24,558 US women without diabetes. In addition to traditional risk factors, the algorithm also includes the emerging risk factor of C-reactive protein (CRP) elevation in its risk calculation. In its initial finding, compared with the FRS-ATP-III, 40-50% of intermediate risk women were reclassified into higher-or-lower risk categories.
In 2008, the RRS for men was similarly developed, using data from 10,724 US men.
American College of Cardiology/American Heart Association Arteriosclerotic Cardiovascular Disease Risk Estimator
The American College of Cardiology (ACC)/American Heart Association(AHA) Arteriosclerotic Cardiovascular Disease (ASCVD) Risk Estimator, released in 2013, was designed to assess the risk of an initial cardiovascular event and includes participants from racially and geographically diverse cohorts such as the Framingham Heart Study (FHS), the Atherosclerosis Risk in Communities (ARIC) study, the Coronary Artery Risk Development in Young Adults (CARDIA), and the Cardiovascular Health Study (CHS).
The pooled cohort equations predict the future risk of cardiovascular disease and also stroke. The variables used were those used in the 10-year Framingham CVD score, but unlike the Framingham CVD, only hard disease endpoints were used in the calculation.
The accuracy of the 2013 ACC/AHA-ASCVD was called into question very soon after its release when investigators calculated the 10-year risk of cardiovascular events in three large-scale primary prevention cohorts—the Women’s Health Study (WHS), the Physicians’ Health Study (PHS), and the Women’s Health Initiative Observational Study (WHI-OS)—and found that the new algorithm overestimated the risk by 75-150% compared with the actual risk in all three cohorts.
Other studies published in 2014 demonstrated increased risk with the ACC/AHA-ASCVD compared with older risk scores. After analysis of data from National Health and Nutrition Examination Surveys (NHANES) from 2005-2010, investigators predicted that use of the ACC/AHA-ASCVD to calculate risk would result in an increase of 12.8 million individuals eligible for statin therapy largely driven by older patients and treating individuals without cardiovascular disease. Among those aged 60 to 75 years, 87.4% of men and 53.6% of women would now be eligible for lipid-lowering medication, which is up from one-third and 21.2% respectively under the older ATP-III guidelines.
Using the NCDR PINNACLE (National Cardiovascular Data Registry Practice Innovation and Clinical Excellence) data from 2008 to 2012, Maddox and colleagues calculated an increase of 32.4% in patients eligible for statin therapy with use of the ACC/AHA-ASCVD.
The QRISK (2007) and the updated QRISK2 (2008) are algorithms developed for the United Kingdom population to better predict 10-year risk of CVD.
In 2009, the UK National Health Service (NHS) began using QRISK2 across its system and since 2010, the UK’s National Institute for Health and Care Excellence (NICE) has recommended QRISK2 as the preferred risk score for assessing cardiovascular risk.
The QRISK2 algorithm includes all the major risk factors of the FRS-ATP-III score, plus the following
Measures of social deprivation
Body mass index (BMI)
Chronic kidney disease
Family history of heart disease
Systematic Coronary Risk Evaluation
The Systematic COronary Risk Evaluation (SCORE) utilized pooled data of over 250,000 individuals from 12 European studies in its development. First published in 2003, the algorithm calculated the 10-year CVD death risk with separate scores for CHD and stroke fatality.
In subsequent revisions, the total CVD risk was also calculated.
Calibrations for high- and low-risk regions of Europe have been developed, based on each country’s mortality statistics. In addition, in the 2012 revision, published in the CVD guidelines released by the Fifth Joint Task Force of the European Society of Cardiology, a cardiovascular risk age calculation was added. The calculation is intended to be used as a communication aid, particularly with younger individuals with low risk estimates.
Comparing Risk Scores
Controversy surrounds the use of CVD risk scores as the basis for guidelines for primary prevention interventions, particularly recommendations for the management of lipid disorders and identifying patients who would benefit from pharmaceutical interventions.
Issues of accuracy of risk calculations are not limited to the ACC/AHA-ASCVD risk estimator. All of the above scores have been validated and perform well in discriminating persons who will develop CVD from those who will not. However, the match between the prediction and actual outcome (score calibration) greatly varies when the algorithms are applied to populations with differing demographics than those of the cohort from which it was developed.
Numerous studies reported calibration disparity in risk assessment using the scores above. For example, a 2015 study utilizing data from the Multi-Ethnic Study of Atherosclerosis (MESA), measured calibration for five risk scores and found the following overestimates for the risk of cardiovascular events
FRS-CHD: 53% in men, 48% in women
FRS-CVD: 37% in men, 8% in women
FRS-ATP-III: 154% in men, 46% in women
ACC/AHA-ASCVD: 86% in men, 67% in women
In this study, the RRS was the best calibrated model, with investigators reporting the lowest discordance between actual and predicted events (-3%).
Two separate studies in 2014 compared the ACC/AHA-ASCVD risk estimator and SCORE in European cohorts. In the first study, which used data from a Dutch cohort 55 years of age and older, the percentages of patients would be eligible for statin therapy, according to the various algorithms, were as follows
ACC/AHA-ASCVD: 96.4% of men and 65.8% of women
SCORE: 66.1% of men and 39.1% of women
FRS-ATP-III 52.0% of men and 35.5% of women
Similarly, in a separate study of a Swiss cohort 50 years and older, the investigators reported that 30 times the number of men were eligible for statin therapy using the ACC/AHA-ASCVD risk estimation as compared with SCORE. In the 60-70 year old age group, twice as many men and six times as many women were considered at high risk.
A 2012 study of men (aged 50-79) and women (aged 47-79) nm ). In general, the RSS predicted increased risk in women and decreased risk in men. The authors concluded that adopting the RRS for the screening of US adults would result in increased clinical management in 1.6 million women and decreased management in 2.10 million men.
A 2014 study comparing the validity of three CVD risk algorithms in a middle-income Asian population in Malaysia found agreement in risk assessment between the US-developed FRS and the European-developed Systematic Coronary Risk Evaluation (SCORE), but the SCORE model was more accurate for predicting risk in men while the FRS was better at predicting risk in women. A third algorithm (World Health Organization/International Society of Hypertension score) performed poorly with both men and women.