Introduction
Despite a 31% decline in the death rate from atherosclerotic cardiovascular disease (ASCVD) (i.e., coronary heart disease [CHD] or stroke) between 2001 and 2011, ASCVD remains the cause of approximately one of three deaths in the United States. More than a third of deaths attributed to ASCVD occur among individuals aged less than 75 years, which is younger than the current life expectancy of 79 years. ASCVD exists on a spectrum, with many high-risk primary prevention patients having ASCVD event rates similar to lower-risk secondary prevention patients (those who have had a prior ASCVD event).
As over half of major ASCVD events occur in previously asymptomatic people, it is critical to identify “at-risk” individuals for early implementation of preventive strategies and treatments. The long incubation period of atherosclerosis allows for such intervention. Indeed, intensive risk factor modification via both lifestyle improvements and pharmacologic treatments of cholesterol, blood pressure, and glycemic control has been shown to modestly regress and stabilize existing atherosclerotic plaques and ultimately reduce ASCVD events.
Unfortunately, unhealthy lifestyle habits and suboptimal risk factor control remain unacceptably high in most populations. Some of the most effective interventions for ASCVD risk reduction are lifestyle modifications. Emphasis must be placed on both preventing ASCVD risk factor development (primordial prevention) and treating existing risk factors (primary prevention).
In 1985, Geoffrey Rose wrote a seminal article entitled “Sick individuals and sick populations.” It conveyed the key message that, despite the fact that high-risk individuals gain the most from preventive measures, the greatest number of deaths from ASCVD occur among individuals at the low- or medium-risk end of the risk distribution, simply because many more people fall into these categories. This became known as the classic Rose paradox and highlights that both “high-risk” and “population-based” preventive strategies are needed and are in fact complementary. It is therefore imperative to employ both population-based and individualized approaches for comprehensive ASCVD prevention ( Fig. 29.1 ). The purpose of individualized risk assessment is to identify persons at a stage when interventions could effectively alter the course of the disease and reduce ASCVD morbidity and/or mortality.
In this chapter, we discuss various tests that can be potentially used for “screening” the asymptomatic individual for the detection of ASCVD. We will distinguish between the roles of traditional population-based screening and individualized risk assessment as strategies for reducing ASCVD risk and optimizing cardiovascular health. There are other tests more commonly used for diagnostic purposes in the evaluation and management of patients with CHD or symptoms suggestive of CHD (e.g., echocardiography, pharmacologic stress testing, coronary computed tomography angiography, cardiac magnetic resonance imaging, and coronary angiography). They are not discussed here in much detail as we will focus on assessment of the asymptomatic individual. Table 29.1 outlines our conceptual distinction between true screening, individualized risk assessment, and diagnostic testing.
| Goal | Target Population | Screening Outcomes | Parameters of Efficacy | Example of Potentially Effective Strategy | Guidelines | |
|---|---|---|---|---|---|---|
| Traditional screening | Early detection of unheralded disease | Healthy people General asymptomatic population | Identify high-risk cases | Sensitivity Specificity PPV NPV | Population-based screening for total cholesterol and blood pressure | USPSTF |
| Individual risk assessment | Individualize risk to inform clinical decision-making | Individual patients “Intermediate” risk Those in whom “treatment decisions uncertain” | Identify lower than expected risk (“derisk”) AND Identify higher than expected risk | C-statistic NRI NNT NNH | Global risk scoring Coronary artery calcium scoring | ACC/AHA ESC/EAS |
| Diagnostic testing | Confirm clinical suspicion, make diagnosis | Symptomatic patients | Make clinical diagnosis | Sensitivity Specificity Negative and positive likelihood ratios | Stress echocardiography | ACC/AHA |
Distinction Between Traditional Screening and Individualized Risk Assessment
Traditional screening is defined as the routine evaluation of a general population with the goal of detection of disease among people without signs or symptoms of the disease, not exclusion of disease. In 1968, Wilson and Jungner outlined 10 criteria for a valid screening program for the World Health Organization (WHO) ( Box 29.1 ) that still hold true today. Briefly, a screening program should be targeted at a disease that is an important health problem, a disease that has a long latent phase where early detection is possible, where treating in an early stage is more beneficial than treating at a later stage, and where potential benefits of screening outweigh the costs.
- 1.
The condition being screened for should be an important health problem.
- 2.
The natural history of the condition should be adequately understood.
- 3.
There should be a latent stage of disease where early detection is possible.
- 4.
Treatment at an early stage should be of more benefit than at a later stage.
- 5.
A suitable test should be devised for the early stage.
- 6.
The test should be acceptable to the population.
- 7.
Intervals for repeating the test should be determined.
- 8.
Adequate health service provision should be made for the extra clinical workload resulting from screening.
- 9.
The risks, both physical and psychological, should be less than the benefits.
- 10.
The costs should be balanced against the benefits.
Thus, screening for ASCVD meets all of these WHO criteria and has great appeal due to the long disease latency, emerging technologies for early detection, and existence of proven therapies that slow its natural history. Nonetheless, various experts have pointed to potential problems with ASCVD screening due to false-positive test results, inappropriate downstream testing, and creation of “pseudodisease.” Pseudodisease means that some individuals with subclinical atherosclerosis may not be destined to have an ASCVD event and there is a risk of overtreatment with medications.
One of the WHO criteria is that costs should be balanced against potential benefits. For ASCVD screening, costs include not only the direct cost of the screening test itself but downstream costs from additional testing, specialty referrals, and treatments; sometimes screening can even lead to more invasive procedures such as coronary angiography and revascularization, which have their own associated risks and expense. In addition, there might be psychological harm from anxiety or being labeled as “diseased” because of subclinical atherosclerosis or mild left ventricular dysfunction.
Several organizations have issued guidelines on ASCVD screening, including the United States Preventive Service Task Force (USPSTF), the American College of Cardiology (ACC)/American Heart Association (AHA), and the European Society of Cardiology (ESC)/European Atherosclerosis Society (EAS). Interestingly, these guidelines often reach different conclusions, owing in large part to different interpretations of the role of screening in routine clinical care. Table 29.2 outlines potential screening or risk assessment tools and the recommendations for or against their use in asymptomatic individuals by these major guideline bodies.
| USPSTF | ACC/AHA | ACC/AHA | ESC/EAS | |
| Grading of recommendations |
|
|
|
|
| USPSTF | ACC/AHA 2010 Risk Assessment | ACC/AHA 2013 Risk Assessment | ESC/EAS | |
| Global risk assessment | 2009 Clinicians should use the Framingham risk model to assess CHD risk and to guide risk-based therapy until further evidence is obtained. | 2010 Global risk scores (such as FRS) should be performed in all asymptomatic adults without known CHD ( class I ). | 2013 Apply race-/sex-specific pooled cohort equations to predict 10-year risk of ASCVD events among non-Hispanic blacks and whites, age 40 to 79 years old ( class I ) and consider using for other race/ethnic populations ( class IIb ). Measure ASCVD risk factors every 4 to 6 years in adults aged 20 to 79 years and estimate 10-year ASCVD risk every 4 to 6 years in adults aged 40 to 79 without known ASCVD ( IIa ). Assess 30-year or lifetime risk in adults aged 20 to 59 without ASCVD who are not at high short-term risk ( IIb ). | 2011 Lipid screening is recommended in those with established ASCVD, diabetes, hypertension, smoking, obesity, family history of ASCVD, family history of dyslipidemias, chronic inflammatory diseases, and chronic kidney disease ( class I ) and in men over age 40 and women over age 50 ( class IIb ). 2012 Global risk estimation using multiple risk factors such as the SCORE estimator is recommended for asymptomatic adults without evidence of ASCVD ( class I, level of evidence [LOE] C ). |
| Electrocardiogram (ECG) at rest | 2012 Recommend against use of a resting ECG for screening asymptomatic adults at low risk for CHD events ( grade D ). Insufficient evidence for the use of a resting ECG for screening asymptomatic adults at intermediate to high risk for CHD events ( grade I ). | 2010 May be reasonable in asymptomatic adults with hypertension and diabetes ( IIa ) or among those without hypertension/diabetes ( IIb ). | 2013 Not addressed. | |
| Treadmill stress ECG | 2012 Recommend against the use of an exercise ECG for screening asymptomatic adults at low risk for CHD events ( grade D ). Insufficient evidence for the use of an exercise ECG for screening asymptomatic adults at intermediate to high risk for CHD events ( grade I ). | 2010 May be reasonable for risk assessment of intermediate-risk asymptomatic adults, especially if non-ECG factors are considered as exercise capacity ( IIb ). | 2013 No recommendation for or against measuring cardiorespiratory fitness. | 2012 Exercise ECG may be considered for ASCVD risk assessment in moderate-risk asymptomatic adults (including sedentary adults considering starting a vigorous exercise program), particularly when attention is paid to non-ECG markers, e.g., cardiorespiratory fitness ( class IIB, LOE B ). |
| Coronary artery calcium (CAC) | 2009 Insufficient evidence for using CAC screening among intermediate-risk adults ( grade I ). | 2010 May be reasonable for risk assessment among intermediate-risk adults (10% to 20% ATP III FRS) ( IIa ) or could be considered for low to intermediate risk (6% to 10%) ( IIb ). | 2013 If risk-based treatment decision is uncertain, CAC ≥ 300 or ≥ 75th age/gender/race percentile may prompt consideration of revising risk assessment upward ( IIb ). | 2012 Measurement of CAC should be considered for ASCVD risk assessment in asymptomatic adults at moderate risk ( class IIa, LOE B ). |
| Carotid intima-media thickness (cIMT) | 2014 Do not screen for carotid stenosis in general population ( grade D ). 2009 Insufficient evidence for using cIMT for risk assessment of intermediate-risk adults ( grade I ). | 2010 May be reasonable for risk assessment among intermediate-risk adults ( IIa ). | 2013 Not recommended ( class III ). | 2012 Measurement of cIMT and/or screening for carotid plaques should be considered for ASCVD risk assessment in asymptomatic adults at moderate risk ( class IIa, LOE B ). |
| Ankle-brachial index (ABI) | 2013 Insufficient evidence to recommend for or against screening in general population ( grade I ); if any benefit, it would be among those at increased risk for peripheral artery disease who are not already receiving interventions for ASCVD risk reduction. 2009 Insufficient evidence for using ABI for risk assessment of intermediate-risk adults ( grade I ). | 2010 May be reasonable for intermediate risk ( IIa ). | 2013 If risk-based treatment decision is uncertain, ABI < 0.9 may promote consideration of revising risk assessment upward ( IIb ). | 2012 Measurement of ABI should be considered for ASCVD risk assessment in the asymptomatic adult at moderate risk ( class IIa, LOE B ). |
| High-sensitivity C-reactive protein (hs-CRP) | 2009 Insufficient evidence for using for risk stratification among intermediate-risk adults ( grade I ). | 2010 Among men ≥ 50 or women ≥ 60 years of age with LDL-C < 130 mg/dL not already on statins (i.e., JUPITER eligibility criteria), may be useful for selecting statin therapy ( IIa ). Among intermediate-risk asymptomatic men < 50 and women < 60 years of age, may be reasonable for risk assessment ( IIb ). Not recommended for high-risk or low-risk adults (men < 50, women < 60 years) ( class III ). | 2013 If risk-based treatment decision is uncertain, hs-CRP ≥ 2 mg/L may promote consideration of revising risk assessment upward ( IIb ). | 2012 hs-CRP may be considered to refine risk assessment in patients with unusual or moderate ASCVD risk profile ( class IIb, LOE B ). hs-CRP should not be measured in asymptomatic low-risk individuals and high-risk individuals to assess 10-year ASCVD risk ( class III, LOE B ). |
| Coronary CT angiography (CTA) | 2010 Not recommended for risk assessment among asymptomatic adults ( class III ). | 2013 Not addressed. | 2013 Not recommended as a screening test in asymptomatic individuals without a clinical suspicion of CHD (class II). |
The USPSTF has consistently advised against routine screening for ASCVD beyond measurement of traditional ASCVD risk factors such as total cholesterol and blood pressure. The recommendations from the USPSTF are best understood in the context of our definition of traditional screening. Because most current tests for the diagnosis of CHD are limited by a low positive predictive value for future CHD events (thus, there is concern for false-positive tests and overtreatment), these tests fail as broad-based population-wide screening tests per the USPSTF criteria.
The ACC/AHA guidelines and the ESC guidelines take a different approach to screening, targeting their recommendations at the individual patient rather than the larger population. This approach is commonly referred to as clinical risk assessment. The goal of individual risk assessment is to inform therapy decisions, particularly in intermediate-risk patients or in patients in whom treatment decisions are otherwise uncertain. As opposed to traditional screening, which focuses exclusively on disease detection, when conducting individual risk assessment, the identification of low ASCVD risk (and, therefore, persons not in need of the aggressive pharmacologic prevention treatment) is equally important. The hallmark of individual risk assessment is that patients may move both up and down the risk spectrum after testing (i.e., risk reclassification). Clinical risk assessment offers the opportunity to rule out high-risk states (and the potential to reduce the intensity of preventive interventions and avoid or withdraw pharmacologic therapy) in a patient who would otherwise have qualified for aggressive treatment based purely on risk factor–based ASCVD screening.
Population-Based Prevention: Is Screening Needed?
Several strategies for ASCVD prevention do not require any type of screening. Foremost among these are purely population-based strategies, which seek to modify risk for everyone across an entire population and are a primary tool for primordial ASCVD prevention. Examples of successful population-based prevention strategies include public smoking bans, trans-fat bans, and salt reduction in packaged and prepared foods.
Some experts have advocated for dismissal of formal screening and adoption of a simple “treat-all” approach to ASCVD prevention. For example, some cost-effectiveness analyses have suggested that treating all adults aged 55 years and over with a low-cost statin may be more cost-effective than any currently available screening strategy. A related strategy adapted to populations with less healthcare resources is the polypill, where all older patients are treated with a combination pill including a statin, aspirin, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker, and thiazide diuretic. Interestingly, the 2013 ACC/AHA Cholesterol Guidelines, which lowered the threshold for recommending statin therapy, have inched closer toward a modified age-based treat-all approach with statin therapy for men over the age of 60 and women over the age of 65. As a general rule, when treatment thresholds are lowered, the importance of traditional population-based screening decreases (because most people will already have qualified for pharmacologic treatment) and the importance of individual risk assessment increases (to determine which individuals might avoid pharmacologic treatment at that point in time and for the next few years).
As a result, in most industrialized countries with greater healthcare budgets, there is an increasing push for greater personalization of preventive therapy. A fundamental goal of individual risk assessment is to limit overtreatment by matching the intensity of preventive therapy to the absolute risk of the patient, thereby maximizing potential benefit while minimizing the potential for harm ( Fig. 29.2 ).
Individualized Screening Starts with Global Risk Assessment
The typical initial approach to screening is ascertainment of traditional ASCVD risk factors, and for specific age-groups (generally excluding the young and the elderly), to estimate one’s 10-year global risk for ASCVD. Multiple risk prediction models have been developed ( Table 29.3 ). Most models include age, sex, smoking status, systolic blood pressure (as well as antihypertensive treatment), total cholesterol, high-density lipoprotein cholesterol, and diabetes mellitus. Some additionally consider family history of ASCVD (Reynolds risk score, QRISK, ASSIGN, PROCAM ) or markers of inflammation (Reynolds risk score), and some of the European risk scoring models (QRISK, ASSIGN) are unique by including measures of social deprivation.
| Region | Variables Included | Outcomes Predicted | Notes on Subgroup | |
|---|---|---|---|---|
| SCORE (2003) | Europe | Age, sex, smoking, SBP, and total cholesterol | 10-year risk of fatal ∗ total ASCVD (MI, stroke, occlusive arterial disease, sudden cardiac death) | Separate equations for low- and high-risk regions of Europe Diabetics already considered high risk |
| Framingham CHD (1998) | Framingham, MA, USA | Age, sex, total cholesterol or LDL-C, HDL-C, SBP, diabetes, smoking | 10-year risk for total CHD (angina, unstable angina, MI, CHD death) | |
| Framingham global CVD (2008) | Framingham, MA, USA | Age, sex, total cholesterol, HDL-C, SBP, BP treatment, diabetes, smoking | 10-year risk of hard ASCVD plus cardiac failure (MI, CHD death, stroke, stroke death, heart failure) | |
| ATP III (2001) | Framingham, MA, USA | Age, sex, total cholesterol, HDL-C, SBP, BP treatment, smoking | 10-year risk of hard CHD (MI or CHD death) | Diabetics already considered high risk |
| Reynolds risk score (2008, men; 2007, women ) | USA | Men: age, total cholesterol, HDL-C, hs-CRP, SBP, smoking, family history of premature CHD Women: age, total cholesterol, HDL-C, hs-CRP, SBP, hemoglobin A 1c , family history of premature CHD | Total CHD including revascularizations (MI, CHD death, stroke, stroke death, coronary revascularizations) | Separate equations for men and women, and for non-Hispanic whites and non-Hispanic blacks |
| QRisk (2007) | United Kingdom | Age, sex, smoking, SBP, BP treatment, ratio of total cholesterol to HDL-C, body mass index, family history of premature CVD, social deprivation | 10-year risk of total CHD including revascularizations (angina, unstable angina, revascularizations, MI, CHD death, stroke, stroke death, TIA) | |
| ACC/AHA pooled cohort equations (2013) | USA | Age, sex, race, total cholesterol, HDL-C, SBP, BP treatment, smoking, diabetes | 10-year risk of hard ASCVD (MI, CHD death, stroke, stroke death) | Races other than white and black should use equation for white race/ethnicity |
| PROCAM (2002) | Germany | Age, LDL-C, HDL-C, SBP, diabetes, smoking, family history of CVD | 10-year risk of hard CHD (MI and CHD death) | Equation for men only |
| ASSIGN (2007) | Scotland | Age, sex, postal code (geography), social deprivation, smoking status and cigarette dosing, family history of CVD, total cholesterol, HDL-C, SBP, diabetes | 10-year ASCVD (CHD, cerebrovascular disease, ASCVD death, revascularization) |
∗ To estimate risk for nonfatal + fatal ASCVD, multiply by 3 in men, by 4 in women, and slightly less for older adults.
The ESC/EAS guidelines recommend using the SCORE system, which estimates 10-year risk of fatal ASCVD, with separate risk estimations for high- and low-risk regions in Europe. In the ESC/EAS guidelines, recommendations for drug treatment of dyslipidemia (vs lifestyle alone) are based on both one’s estimated 10-year risk of fatal ASCVD using SCORE (with < 1% being low risk, 1–4% moderate risk, 5–10% high risk, and ≥ 10% very high risk) and one’s low-density lipoprotein cholesterol (LDL-C) level. In general, individuals at higher predicted fatal ASCVD risk are recommended for drug treatment at much lower LDL-C levels.
In the United States, the 2013 ACC/AHA guideline on the assessment of cardiovascular risk endorses risk factor screening every 4 to 6 years for those aged 20 to 79 years and application of the pooled cohort equations in asymptomatic adults aged 40 to 79 years to estimate 10-year risk for a first “hard” ASCVD event (myocardial infarction [MI] or stroke). The ACC/AHA risk estimator was not designed to be used in those already on statin therapy. In contrast to previous cardiovascular risk scores, there are now separate models by race (e.g., non-Hispanic whites and blacks) and by sex for more refined risk prediction. These guidelines are linked to the ACC/AHA cholesterol guidelines, which state that higher-risk individuals (10-year predicted risk of ASCVD ≥ 7.5%) are recommended for statin treatment, after a clinician-patient discussion. Moreover, those persons with a 5% to 7.4% risk over the next decade can also be considered for moderate-intensity statin therapy after a clinician-patient discussion. In addition, the ACC/AHA guidelines support estimating a 30-year or lifetime ASCVD risk based on traditional risk factors for adults aged 20 to 59 who are not at high short-term (i.e., 10-year) risk.
The USPSTF also recently posed draft recommendations on statin use for the primary prevention of ASCVD in adults on its website. They recommended that the people who benefit most from statin use are 40 to 75 years of age with at least one other risk factor for ASCVD and have a 10-year ASCVD risk estimate of 10% or greater (grade B recommendation). They advise that those with a 7.5% to 10% estimated 10-year risk are less likely to benefit and should talk to their doctor (grade C recommendation).
Risk Uncertainty Often Remains After Global Risk Assessment
In contrast to the ASCVD risk estimation tools developed for the United States and Europe, there are many areas of the world such as the BRICS countries (Brazil, Russia, India, China, and South Africa) where accurate tools have not been developed; yet these countries have a large burden of cardiovascular disease. Current risk estimation calculators may over- or underestimate risk in these populations. Caution should be used when applying the ACC/AHA pooled cohort equations to groups outside the United States and to groups that are neither white nor black within the United States. This may result in overestimation of ASCVD risk among those of Chinese/East Asian descent and underestimation in American Indians and those of South Asian descent.
In addition, in the context of epidemiologic trends pointing toward continually decreasing ASCVD rates in the United States and in many developed countries, there remains a concern that risk scores such as the ACC/AHA pooled cohort equations that are based on historic data may systematically overestimate ASCVD risk in modern populations.
Furthermore, with all risk calculators, applying the results from population-based estimators to clinical decision-making at the individual level can be problematic. Specifically, these calculators estimate the average risk in a group of individuals who have similar risk factor profiles; however, a given risk score is far more accurate for this group than it is for any individual within the group. Indeed, with respect to the individual, each estimate has a theoretical confidence interval that is unknown and that may overlap personalized treatment thresholds.
A critical feature of the 2013 ACC/AHA guidelines is the acknowledgment that in many cases after initial 10-year risk estimation, the treatment decision about statin therapy will still remain uncertain to either the patient or the clinician. The guidelines allow for revising one’s risk status upward if one of the following is present: elevated lifetime risk, family history of premature ASCVD, high-sensitivity C-reactive protein (hs-CRP) ≥ 2.0 mg/L, LDL-C ≥ 160 mg/dL, abnormal coronary artery calcium (CAC) score (Agatston score ≥ 300 or ≥ 75th percentile for one’s age and sex), or ankle-brachial index (ABI) < 0.9.
The ESC/EAS guidelines, which endorsed the SCORE estimator as a starting point, also acknowledged uncertainty by including many important “qualifiers” in their document where risk estimation may need to be adjusted upward or downward based on an individual’s pretest probability. These guidelines state that (1) risk will be overestimated in countries with falling ASCVD mortality rates and underestimated in countries where mortality is increasing; (2) at a given age, women will have a lower predicted 10-year ASCVD risk than men, but this may be misleading because eventually similar numbers of women will die of ASCVD compared to men; (3) risk will be higher in individuals with social deprivation; (4) risk will be higher among those who are sedentary and have central obesity; (5) risk will be higher among those with low high-density lipoprotein (HDL)-C, increased triglycerides, increased apolipoprotein B, or increased hs-CRP; (6) risk will be higher among asymptomatic individuals with evidence of subclinical atherosclerosis; (7) risk will be higher among those with impaired renal function; (8) risk will be higher among those with a family history of premature ASCVD; and (9) risk will be lower among those with high HDL-C or a family history of longevity.
Despite these limitations, global risk estimation tools are helpful in initiating a risk-based discussion of preventive pharmacologic therapy, and we strongly endorse their routine use as a starting point. But, given these caveats of risk underestimation among certain patient subgroups (i.e., potential for undertreatment) alongside concerns for risk overestimation in other populations combined with much lower treatment thresholds (i.e., potential for overtreatment), the need for further individualized risk assessment has become increasingly important.
With their extensive safety data and the availability of low-cost generic statins, combined with much lower treatment thresholds, many more patients are now eligible for statin therapy under the 2013 ACC/AHA guidelines. The group previously recommended for risk reclassification (the “intermediate-risk” group) has become much more narrow and with lower absolute risk; however, uncertainty of risk often remains among a broad range of estimated risk scores (5–15% 10-year risk). We discuss later in this chapter how selective use of additional tools may help refine personalization of risk assessment. Select tests are discussed in detail hereafter, with a particular emphasis on exercise stress testing (to assess cardiorespiratory fitness or exercise capacity) and noncontrast cardiac computed tomography (CT) (to measure the presence of or the amount of CAC), which appear to be the best predictors of long-term survival.
Additional Medical and Social Factors for Revising Risk Estimation Upward
Before one considers ordering additional testing to refine risk, there are additional elements that can be obtained from a detailed medical and social history that help guide risk assessment. For example, there are many risk factors for ASCVD that are not included in the pooled cohort equations including a family history of premature or later-onset ASCVD, former cigarette smoking, secondhand exposure to smoke, history of erectile dysfunction (ED), history of adverse pregnancy outcomes (i.e., preeclampsia or gestational diabetes), insulin resistance/prediabetes/metabolic syndrome, sedentary behavior, autoimmune disease, human immunodeficiency virus infection, chronic kidney disease, obstructive sleep apnea, and hepatosteatosis. Several of these are discussed herafter.
It would be cumbersome and impractical to add these and other unique risk factors to a universal risk prediction model. However, physicians require further direction on how best to categorize individuals with unique risk factors. It is especially in this setting that further risk stratification with tests that inherently individualize risk, such as measuring subclinical atherosclerosis, may better delineate those who would most likely benefit from preventive therapies.
Family History of Premature Coronary Heart Disease
Screening for ASCVD in the asymptomatic individual should incorporate a detailed family history. A history of premature CHD among first-degree relatives has been shown in multiple epidemiologic studies to be strongly associated with incident ASCVD including MI, coronary death, and stroke. Studies have demonstrated a 2- to 7-fold increase in risk of CHD associated with a positive family history, with men younger than 60 years of age being the most affected.
Family history, which is asked by virtually every clinician conducting clinical risk assessments, was considered for inclusion in the 2013 ACC/AHA risk models. However, it was not included as part of the ACC/AHA pooled cohort equations because it did not adequately improve model performance, likely because it was not distinguished from a family history of premature CHD—a well-established predictor of subsequent ASCVD events. On the other hand, other risk score calculators (Reynolds risk score, QRISK, ASSIGN, PROCAM) do include family history of premature ASCVD in risk assessment.
However, the ACC/AHA risk assessment guidelines do state that the presence of a family history of premature CHD (described as occurring in male first-degree relatives before age 55 and female first-degree relatives before age 65) could be used to revise one’s risk estimation upward (IIb recommendation, level of evidence B). A family history of premature CHD is similarly endorsed as a significant risk factor by experts in the 2011 AHA effectiveness-based guidelines for the prevention of cardiovascular disease in women. Additionally, the Canadian Cardiovascular Society guidelines recommended that a person’s estimated risk be doubled with a family history of premature ASCVD. A family history of premature CHD in a subject’s parents is one of the factors included in the Reynolds risk score for ASCVD prediction. A history of ASCVD in a sibling is an even stronger risk factor than a parental history.
However, not everyone with a positive family history of premature CHD is destined to have an ASCVD event, and thus upgrading the risk category (and potential eligibility for statins) for everyone with a family history may be inappropriate. Moreover, many individuals with a family history of premature CHD have few or no other risk factors, and it may be difficult to accurately determine their risk. When risk is uncertain, selective use of subclinical atherosclerosis imaging tools such as CAC may help guide the risk discussion.
Autoimmune Diseases
Autoimmune diseases, such as systemic lupus erythematous, scleroderma, psoriatic arthritis, and rheumatoid arthritis, affect approximately 8% of the population, 78% of whom are women. Inflammation underlies the development of atherosclerosis, and autoimmune rheumatic diseases are associated with higher rates of cardiovascular morbidity and mortality due to accelerated atherosclerosis. Multiple studies have demonstrated the association between rheumatoid arthritis or systemic lupus erythematous and increased risk for ASCVD. Thus, patients with autoimmune diseases may warrant more intensive preventive therapies or may be candidates for additional second-tier risk assessment tools such as CAC, if risk is uncertain. Additionally, treatment with biologic agents, such as antitumor necrosis factor agents, has been shown to decrease the risk for cardiovascular events in rheumatoid arthritis patients.
Adverse Pregnancy Outcomes
Pregnancy complications, such as gestational diabetes and preeclampsia, give insight into a mother’s long-term ASCVD risk. Thus women (even years past their child-bearing days) should be asked about their prior pregnancy outcomes as part of an office-based ASCVD risk assessment.
Gestational diabetes is associated with long-term adverse maternal ASCVD risks, such as type 2 diabetes mellitus, hypertension, and metabolic syndrome. Nearly half of women with a history of gestational diabetes mellitus will develop type 2 diabetes mellitus within 10 years. However, gestational diabetes is also a risk factor for the development of ASCVD independent of conventional risk factors, especially among women with elevated body mass index.
Preeclampsia is a multisystem disease that occurs after 20 weeks of gestation, mediated by abnormalities in the placental vasculature leading to both short- and long-term endothelial dysfunction and inappropriate vasoconstriction in multiple vascular beds. It presents with hypertension and proteinuria and complicates approximately 2% to 8% of pregnancies. Some risk factors such as diabetes and obesity may predispose women both to preeclampsia during child-bearing age and also to increased ASCVD risk later in life (i.e., the pregnancy “unmasks” underlying predisposition to atherosclerotic vascular disease). Alternatively, preeclampsia may directly have a causal effect on the vasculature that contributes to ASCVD later in life.
Preeclampsia is independently associated with an increased risk for ASCVD events. Women with a history of preeclampsia have nearly double the risk of CHD and stroke approximately 10 to12 years later and a nearly 50% increased risk of all-cause mortality at an average of 15 years’ follow-up. Thus, the 2011 AHA women’s prevention guidelines consider these adverse pregnancy outcomes to be significant risk factors for ASCVD —on par with traditional risk factors such as smoking and hypertension.
Erectile Dysfunction
ED is very common, affecting 40% of men older than 40 and 70% of those over 70 years of age. Up to 80% of ED is from vascular etiologies. Data from the National Health and Nutrition Examination Survey (NHANES) found that almost 90% of men with ED had at least one major ASCVD risk factor (hypertension, hypercholesterolemia, current smoking, or diabetes). A 2013 Australian study of 95,000 men with no known heart problems found that those with severe ED had a 60% higher risk of developing heart disease and nearly twice the risk of dying, compared with those without ED. Other studies have reaffirmed these findings.
ED may be a marker of generalized endothelial dysfunction. Because ED symptoms can precede ASCVD, screening for ED can be used as an early marker to identify men at higher risk of ASCVD who might benefit from intensive treatment of risk factors and a detailed cardiovascular assessment. More than 40% of men with ED and risk factors for ASCVD are unaware of their risk. The ESC guidelines state that all men with ED should undergo ASCVD risk estimation and risk management (class IIa recommendation).
A 2015 study determined that screening men with ED for ASCVD risk would be a cost-effective strategy that would not only help avert ASCVD events but also potentially save more than $21 billion dollars in healthcare expenses over 20 years in the United States alone. When ASCVD risk is uncertain, the use of personalized prognostic tools, such as the CAC score, may aid in cardiovascular risk stratification and management of men with vascular ED.
Assessment of Physical Inactivity and Sedentary Behavior
The AHA recommends that assessment of physical activity levels should be a vital health measure that is screened for at regular intervals similar to all other major modifiable ASCVD risk factors (i.e., diabetes mellitus, hypertension, hyperlipidemia, obesity, smoking) and tracked over time. More than half of the American population perform less than the recommended levels of physical activity, and annually approximately 250,000 deaths in the United States can be attributed to the consequences of physical inactivity. Furthermore, from a global perspective it is estimated that 31% of the world’s population are not getting recommended levels of physical activity, and physical inactivity has been cited as the fourth leading cause of death worldwide.
The AHA has released a document establishing goals to improve the cardiovascular health of all Americans by 20%, while reducing deaths from ASCVD by 20% before 2020. In this statement, ideal cardiovascular health is defined by seven factors, one of which is performing recommended levels of physical activity. Similarly, WHO and other global advocacy groups have listed physical activity as a public health priority.
Recent data have highlighted the importance of sedentary behavior as a distinct risk factor. Whereas physical inactivity is the failure to meet the recommended moderate to vigorous physical activity threshold, sedentary behavior refers to the time spent in behaviors that result in ≤ 1.5 metabolic equivalents of task (METs). This commonly includes any seated or reclined posture at a desk, in a car or bus, using a computer, or watching television and generally excludes time sleeping. Therefore, even individuals who exercise on a treadmill each morning can have prolonged sedentary time if much of their day is spent working at a desk, driving in a car, and relaxing at home in the evening. Biswas et al. showed that prolonged sedentary time is independently associated with ASCVD events, cardiovascular mortality, all-cause mortality, and other adverse health outcomes.
The study of Biswas et al. and much of the physical inactivity and sedentary behavior research literature have a limitation in that they rely on self-report and use of surrogate measures such as time spent watching television. Lack of access to more objective data on behavior is also a challenge in medical practice. One option to consider using is the standard pedometer in conjunction with a step count diary. In addition, new mobile health (mHealth) technologies, namely built-in smartphone activity trackers and wearable connected health devices, equipped with triaxial accelerometry, are increasing in popularity and appear accurate. These devices may facilitate even easier access to longitudinal information on step counts, the time spent in various forms of activity, and the pattern of activity. They likely will enhance the quality of future research in this field, and, when data are integrated with the electronic medical record, this could facilitate implementation of physical activity/sedentary behavior as a vital sign in clinic. Moreover, a digital data stream could enable automated interventions providing real-time coaching to increase physical activity and reduce sedentary behavior.
Chronic Kidney Disease
Chronic kidney disease (CKD) is an independent risk factor for ASCVD, and the risk of ASCVD among CKD patients is not fully captured by traditional risk factor models such as the Framingham risk score or the 2013 ACC/AHA pooled cohort equations. More than 50% of deaths in CKD patients are attributed to ASCVD, and CKD patients are more likely to die from ASCVD than to progress to end-stage renal disease. Patients with CKD should be assessed for ASCVD risk, and they may benefit from more intensive preventive therapies similar to those for patients with established CHD. The 2011 AHA women’s prevention guidelines consider CKD a high-risk condition, similar to clinically manifest ASCVD and diabetes. The ESC guidelines also give a class I recommendation that patients with CKD should have risk factors managed in a similar way as individuals with very high ASCVD risk. In the REasons for Geographic and Racial Differences in Stroke (REGARDS) study (a natural history study), only approximately 8% of patients with CKD in the 50 to 79 years age range did not meet 2013 ACC/AHA cholesterol guideline criteria for initiating a statin regimen.
Metabolic Syndrome
The metabolic syndrome, a constellation of three or more of abdominal obesity, prehypertension, prediabetes, low HDL-C, and high triglycerides, is an important marker of risk for ASCVD and new-onset type 2 diabetes, largely due to the influence of abdominal obesity and insulin resistance. Although not itself a risk-scoring tool, the metabolic syndrome clinical construct is meant to draw attention to the clustering of cardiometabolic risk factors in certain predisposed individuals. A clinical diagnosis of metabolic syndrome can help facilitate communication with patients and other providers about the core causative factors of a patients ASCVD risk, the risk of developing diabetes, and importance of lifestyle modification.
There are several ancillary hallmarks of metabolic syndrome that can be important for identifying ASCVD risk. For example, the most common liver function abnormality in the United States is asymptomatic fatty liver disease (hepatosteatosis), which is a core component of the pathophysiology of atherogenic dyslipidemia in metabolic syndrome. Obstructive sleep apnea is also commonly seen among patients with the metabolic syndrome and is associated with ASCVD risk beyond what would be predicted by traditional risk factors alone. The ESC guidelines state that all individuals with obstructive sleep apnea should undergo ASCVD risk assessment and risk factor management.
Predisposition to metabolic syndrome is also one of the reasons why patients of South Asian ancestry appear to have elevated ASCVD risk beyond what would be predicted by traditional risk factors alone. Other patients predisposed to metabolic syndrome include patients with human immunodeficiency virus, patients with depression, patients treated with antipsychotics, treated cancer survivors, and patients with prior organ transplants.
Additional Tools for Screening or Refining Individualized Risk Assessment
Resting Electrocardiogram
One of the simplest potential screening tests for ASCVD is a resting electrocardiogram (ECG); this is a low-cost test, with essentially no direct risk from the test itself, and it is readily available in most clinical practices. Major ECG abnormalities include Q waves suggestive of silent MI, evidence of left ventricular hypertrophy (LVH), complete bundle branch block, atrial fibrillation or flutter, or major ST-T wave changes. Minor ECG abnormalities include minor ST-T changes.
In several population-based studies of asymptomatic adults, ECG abnormalities predicted incident CHD events and mortality. In a population based study of older adults (aged 70 to 79 years), both major and minor ECG abnormalities were associated with increased risk of CHD events when added to a model with traditional risk factors (although not when added to the Framingham risk score); the largest reclassification was seen for the intermediate-risk group (14% reclassified). Of note, the main contribution to the net reclassification improvement in this study came from reclassifying patients to lower, not higher, risk, which actually would result in fewer patients receiving preventive pharmacologic treatment; if this downward reclassification was not appropriate, this could adversely lead to fewer ASCVD events prevented. A resting ECG is not a sensitive enough tool to exclude the presence of moderate or advanced coronary artery disease (CAD).
Although not endorsed in the 2013 ACC/AHA risk assessment guidelines, the 2010 ACC/AHA guideline for the assessment of cardiovascular risk in asymptomatic adults stated that an ECG at rest may be reasonable in asymptomatic adults with hypertension or diabetes (class IIa recommendation). They also gave a weaker but still generally supportive recommendation even among those without hypertension or diabetes (class IIb).
In contrast, based on a systematic review, the USPSTF did not recommend screening asymptomatic adults at low CHD risk with a resting or exercise ECG for the prediction of CHD events (grade D recommendation). Similarly, this same sentiment was also endorsed in a 2015 statement by the American College of Physicians, which stated that clinicians should not screen asymptomatic low-risk adults with a resting or exercise ECG for the detection of cardiac disease given low pretest probability and low likelihood that positive results will affect treatment decisions and clinical outcomes, plus the potential harms of false-positives leading to unnecessary tests and procedures.
Resting Echocardiogram
A resting transthoracic echocardiogram is widely used to assess patients with suspected cardiac symptoms or structural heart disease, both for diagnostic purposes and management. However, it is not recommended for routine use for screening of asymptomatic individuals.
The majority of echocardiograms are ordered by primary care physicians, rather than cardiologists. Population-based studies of asymptomatic individuals screened by echocardiography have found that incidental findings such as asymptomatic left ventricular (LV) dysfunction and LVH can predict cardiovascular and all-cause mortality independent of blood pressure and other risk factors. The 2010 ACC/AHA guidelines for assessment of cardiovascular risk in asymptomatic adults provide weak support for selective use of echocardiography screening to detect LVH and LV dysfunction for asymptomatic adults with hypertension (class IIb recommendation) but a class III (no benefit) for those without hypertension.
Of note, the population-based Tromsø study in Norway evaluated outcomes among asymptomatic participants who were randomly assigned to a screening echocardiogram or to a control group. Participants with abnormal findings on echocardiogram newly detected by screening (including myxoma, LV dysfunction, wall motion abnormality, or valvular disease) were referred to a cardiologist for further evaluation (approximately 9% of participants). However, despite cardiology referral for these diagnoses, ASCVD events and all-cause mortality during 15 years of follow-up were unchanged between the screened and nonscreened groups.
In response to the growing use of echocardiography (∼8% annual increase ), in 2011 the American Society of Echocardiography (ASE) updated its appropriate use criteria consensus statement in conjunction with the ACC and AHA. This document reaffirmed that echocardiography should not be used for routine screening in an unselected general population, including patients with asymptomatic hypertension. The consensus statement also does not endorse screening asymptomatic family members of individuals with ASCVD with echocardiography unless there is a first-degree relative with inherited cardiomyopathy or suspected connective tissue disease.
Although echocardiography does not use any ionizing radiation, a normal resting echocardiogram does not exclude risk for significant CAD. Therefore, patients with a “normal” appearing echocardiogram may be falsely reassured and not follow through with other recommended screening or preventive measures. Mild abnormalities or questionable test results may lead to additional testing associated with expense and potential for harm.
Exercise Treadmill Testing
Some patients and providers may equate screening for CHD with exercise treadmill testing (ETT). However, ETT can generally only detect obstructive CHD, whereas the majority of MIs occur from acute plaque rupture from thin-capped fibroatheromas producing < 50% diameter stenosis. Whereas ETT is commonly used to evaluate symptoms suggestive of possible CHD for diagnostic purposes, current guidelines do not support routine ETT in the vast majority of asymptomatic individuals.
For certain subgroups of asymptomatic patients, however, ETT may be useful. The previous 2010 ACC/AHA guidelines gave a modest class IIb recommendation that ETT may be considered for ASCVD risk stratification for the intermediate-risk individual, particularly when non-ECG parameters such as exercise capacity (i.e., METS achieved) are considered. Other older ACC/AHA guidelines felt it was reasonable (class IIa recommendation) to screen asymptomatic diabetic individuals who plan to start vigorous exercise but gave a weak endorsement (class IIb) for older nondiabetic individuals before vigorous exercise. Similarly the ESC stated in their guidelines that exercise ECG may be considered for ASCVD risk assessment in moderate-risk asymptomatic adults (including sedentary adults considering starting a vigorous exercise program), particularly when attention is paid to non-ECG markers like cardiorespiratory fitness (class IIb).
However, in contrast to the older ACC/AHA and ESC guidelines, given the low prevalence of significant obstructive CAD among asymptomatic individuals screened, the USPSTF determined there was insufficient evidence for routine stress testing before exercise training and did not endorse screening.
Unfortunately, no randomized controlled trials (RCTs) have addressed the utility of ETT in asymptomatic individuals, even those pursuing a vigorous exercise program or competition. Further discussion regarding screening of competitive athletes is found in the “special populations” section of this chapter.
The role of ETT screening for high-risk occupations like commercial and military pilots and competitive athletes also remains highly controversial, but current evidence does not support the role of routine screening among these individuals either. A study analyzing the use of ETT screening of asymptomatic US Air Force pilots found a very low positive predictive value of CAD of only 16%, and thus screening ETT was not felt to be efficacious.
An abnormal ETT is associated with an increased risk of MI and sudden cardiac death, though the positive predictive value is low. Many parameters obtained by ETT individually have prognostic value, and they include ST-segment changes, assessment of exercise capacity, heart rate recovery, arrhythmias, and blood pressure response. Reduced exercise capacity (METs) and poor heart rate recovery are strongly predictive of ASCVD death even independent of traditional ASCVD risk factors. Conversely, those with excellent exercise capacity generally have a favorable long-term prognosis.
Exercise capacity/physical fitness is one of the strongest overall predictors of survival. The Henry Ford Exercise Testing Project (FIT) evaluated over 58,000 patients without known CHD who underwent a clinically indicated ETT and then were followed for an average of 10 years for mortality and CHD outcomes. After age and sex, the factors most predictive of survival were METs achieved and percentage of maximal predicted heart rate achieved. This FIT treadmill score, using standard parameters obtained from ETT, as well as age and sex, can be used to estimate one’s 10-year mortality risk and potentially refine ASCVD risk prediction. This score is defined as percentage of maximum predicted heart rate + 12 (METs) − 4 (age) + 43 if female. For prognostic purposes, the FIT treadmill score, which runs from −200 to 200, can be considered to be included in all ETT reports for patients who have no existing ASCVD and have a negative ECG portion on their stress test.
In addition to being inversely associated with CHD events and mortality, greater fitness (assessed by METs) is also associated with decreased risk of other important outcomes such as atrial fibrillation, incident diabetes, incident hypertension, and heart failure. Importantly, high fitness/exercise capacity achieved antecedent of a first heart attack is perhaps the best predictor of heart attack survival. The good news is that only moderate fitness levels are required to significantly improve the coronary risk factor profile.
As previously discussed, greater levels of physical activity are strongly associated with reduced ASCVD risk. Because most asymptomatic patients do not undergo an ETT for risk stratification, asking about physical activity is frequently used as a surrogate for fitness in clinical practice. However self-reported physical activity is only modestly correlated with directly measured fitness and when discordant, measured fitness is a better marker of cardiometabolic risk.
Despite the strong prognostic value of fitness measures for identifying individuals at increased risk for ASCVD outcomes, it remains unclear how to use the ETT risk-stratification results to alter the management of patients compared to current guideline recommendations. Achievement of a high MET level (generally > stage 4 on the Bruce protocol) and high FIT score confer excellent prognostic information; such individuals may generally be reassured and recommended to continue following a healthy active lifestyle. However, less favorable findings from an ETT could lead toward more targeted and intensified efforts at promoting improved physical activity and risk factor control for at-risk individuals (i.e., aggressive lipid-lowering therapy ).
Coronary Artery Calcium
ASCVD risk prediction models are heavily weighted toward chronologic age, yet “arterial” or “biologic age” is frequently discordant with chronologic age. Assessment of CAC is a useful surrogate measure of total coronary atherosclerotic burden and therefore arterial age.
CAC is performed using noncontrast cardiac-gated CT scanning. CAC scans can be performed on any modern CT equipment and thus can be performed similarly around the world. The entire procedure takes about 10 minutes, with most of the time spent placing electrodes and positioning the patient on the table. With modern scanners, scans are acquired in under a second using approximately 0.5 to 2 mSv of radiation (approximately equivalent to 2 bilateral mammograms or 10 chest x-rays). In most metropolitan centers in the United States, CAC scans cost between $75 and $150.
CAC scanning leverages the fact that calcified deposits in the coronary arteries strongly attenuate x-rays and are thus visible on unenhanced images. Contiguous voxels approximately 130 Hounsfield units are considered to be calcium. CAC scans are generally scored using the Agatston score, which is a summed score of all calcified lesions in the coronary arteries through the complete z-axis of the heart, weighted for the density (x-ray attenuation) of the calcium. Agatston scoring uses a 120-kV electron, variable mA based on patient body weight, acquiring up to 40 slices at a fixed 2.5-mm to 3-mm slice increment. The ideal way to score a CAC scan remains controversial, with potential benefit in differently accounting for the density of calcium and regional distribution of calcium, as well as extracoronary calcification. Future advancements will likely increase the sensitivity to minute calcium deposits and will likely further reduce the associated radiation.
CAC scoring does not identify isolated noncalcified plaque, although its prevalence is less common when calcium is not present anywhere in the coronary tree. Current evidence does not support a strong incremental predictive value for the detection of isolated noncalcified plaque in the asymptomatic primary prevention patient.
When added to traditional risk factor models, CAC improves risk discrimination more than any other available test. In the community-based Multi-Ethnic Study of Atherosclerosis (MESA), the area under the receiver operating characteristic (ROC) curve for the prediction of CHD events was significantly improved when a CAC-based model was compared to a model with only traditional risk factors. Individuals without risk factors but elevated CAC have substantially higher event rates than those who have multiple risk factors but no CAC. In MESA, individuals with no risk factors and CAC greater than 300 had an event rate 3.5 times higher than individuals with 3 or more risk factors and CAC of 0. Even minimal CAC (scores 1 to 10) is associated with 3-fold increased CHD risk compared to those with a CAC of 0. Additionally, there appears to be no upper threshold of risk with increasing CAC scores.
In addition to its role in “upgrading” or elevating predicted risk in younger patients when significant CAC is present, perhaps the most important potential role of CAC testing in the modern era using currently available risk scores may be for downgrading or “derisking” an older adult with a CAC score of 0 who might otherwise be recommended for pharmacologic therapy based on models that are dominated by chronologic age. We have previously found that the number needed to treat to prevent one ASCVD event would generally be unfavorable using a high- or moderate-intensity statin, a polypill, or aspirin for primary prevention when CAC = 0 ( Table 29.4 ).
