Chapter 1: Introduction

2.1 Coronary morbidity

In the hospital different labels are used to classify patients presenting with chest pain, dyspnoea, syncope or other cardiac symptoms.

The widespread introduction of troponins into the diagnostic processes is likely to lead to a shift in the proportion of patients from those formerly diagnosed with unstable angina to those diagnosed with non-Q-wave myocardial infarction [41].

Stable angina is a well-recognized clinical syndrome whose natural history is well described. The incidence of new cases of stable angina pectoris diagnosed in primary care is about 2 per 100 per annum among patients aged over 45 years in Finland [42], considerably higher than the incidence of acute myocardial infarction. Intriguingly neither the incidence of diagnosed angina nor the prevalence of typical angina symptoms in the general population [43] show the marked male excess characteristic of myocardial infarction. Stable angina pectoris confers a markedly increased risk of coronary death compared to expected rates in the general population. In relative risk terms these effects are as strong in women as they are in men, based on a large series of primary care diagnosed patients in Finland [42]. The absolute risk of coronary death and nonfatal myocardial infarction among patients with angina (but without a history of heart attack) depends on case definition. In this national sample aged 45-89 years, about three quarters of all cases were treated pragmatically, without diagnostic test abnormality, among whom annual event rates were 1.28 per 100 person years in women and 2.76 in men. Among patients with an abnormal coronary angiogram or exercise ECG test the annual event rates were higher at 3.74 and 6.51, respectively. Thus patients with angina constitute a ‘high’ risk group, even among women and those without test abnormalities.

As with other manifestations of chronic ischaemic heart disease the burden on the health services in terms of diagnostic and revascularization procedures is predicted to rise in line with an increase in the proportion of elderly in European populations.

2.2 Heart failure

Pump failure of the heart is a common cause of death in the elderly although this not always reflected in mortality statistics because of the limitations of coding rules. Hospital admission rates for heart failure have been increasing in the United States [44] and in Europe [45–47]. The epidemiology of heart failure is described elsewhere [48, 49]. Hypertension, obesity and diabetes are major risk factors. Although a small proportion of clinical cases are due to valve disease (often linked with ischaemia), or to cardiomyopathy, epidemiological studies suggest that, in well developed countries, the majority of cases are due to ischaemia [50, 51].

2.3 Aortic aneurysm and dissection

Aortic aneurysm is also atherosclerotic in nature and increasing mortality trends have been shown in some European countries [52]. It is a potentially preventable cause of death, particularly when confined to the abdominal aorta. The prevalence is 5% in men aged 60 years or more and 1-2% in women. Screening for this condition has been suggested since elective surgical repair carries a 5-8% 30-day mortality in comparison with 50% mortality for ruptured aneurysm; a trial of screening conducted in the United Kingdom has shown encouraging results [53].

2.4 Peripheral arterial disease

It is known that coronary and peripheral vessels are affected by the same disease process, requiring the same treatment modalities. However, despite the high prevalence, until recently little attention has focused on the prevention and treatment of peripheral arterial disease (PAD) or ‘arteriosclerosis obliterans'. PAD occurs almost as frequently in women as in men [54]. The disease is often asymptomatic and underdiagnosed. Over half of the patients are asymptomatic and of patients with mild to moderate PAD who are symptomatic, only about one third report intermittent claudication [55]. Patients with PAD have much higher rates not only of limb amputation but also of coronary heart disease, myocardial infarction, stroke and death [56]. Despite the well known strong association with cardiovascular morbidity and mortality, patients with PAD are less likely to receive appropriate treatment for their risk factors than are those being treated for CHD. The correlation of PAD with CHD reflects the widespread nature of atherosclerosis. However, some minor differences have emerged from epidemiological studies regarding the risk factors for these diseases. Smoking appears to be more important in the aetiology of PAD than in CHD [57]. A positive family history, hypertension, diabetes, dyslipidemia including increased total and LDL-cholesterol and decreased HDL-cholesterol, increased fibrinogen and C-reactive protein, advanced age and physical inactivity seem to be common risk factors. The role of increased Lp(a) and homocysteine in PAD still remains unclear [58, 59]. The same is true for endogenous sex hormones [60, 61]. As with CHD, prevention is essential. Risk reduction can be achieved through lifestyle modification, particularly physical activity and exercise therapies, smoking cessation, use of statins, antiplatelet therapies, antithrombotic strategies, ACE inhibitors and β-blockers [54, 55, 62].

The beneficial effects of statins in these patients have been shown in large trials [63]. Statins not only lower the risk of PAD and vascular events, but they also improve the symptoms associated with PAD. There is also evidence that statins reduce surgical mortality and improve graft patency and limb salvage in PAD patients [64]. Statin therapy also improves cardiovascular outcomes of patients with PAD [65].

2.5 Stroke

In these guidelines, the emphasis is on ischaemic stroke while intracerebral haemorrhage and subarachnoid haemorrhage (SAH), which contribute 10 and 5% of strokes, respectively, are dealt with more cursorily. However, haemorrhagic strokes are often included in the term ‘stroke’ in epidemiologic studies.

The average age of patients in population-based studies of first stroke is usually 70 to 76 years, which is approximately 10 years older than patients with MI. A few studies have shown a decreasing trend in stroke incidence over the last decade, while others have found stable or increasing trends [66].

Stroke is defined as rapidly developed clinical signs of focal (or global) disturbance of cerebral function, lasting more than 24 h with no apparent cause other than a vascular origin. If symptoms last less than 24 h the attack is called a transient ischaemic attack (TIA) [67–69]. Distinction between TIA and ischaemic stroke is arbitrary. In TIAs lasting more than 1 h cerebral infarction may be found in about 20% of cases when examined by MRI. TIAs may precede clinical stroke and require the same prophylactic treatment as ischaemic stroke.

The incidence of stroke increases exponentially with age affecting about 25 per 100 000 in the age group 35-44 years and 1500 per 100 000 in age group 75-84 years per year. It is projected that even with stable age-specific incidence rates, the total number of strokes will increase in the coming decades due to increased life expectancy [66]. Therefore, measures to prevent stroke are needed.

Ischaemic stroke may be due to large vessel disease, small vessel disease, emboli from the heart or from the aortic arch, or other rarer identified causes while a large proportion still remains undetermined [70]. The pathophysiology in large vessel disease of carotid and vertebro-basilar arteries resembles the lesions in the coronary arteries, while that of the other causes of ischaemic stroke differs. Even in the case of large vessel disease the damage is more frequently caused by artery-to-artery embolism than to local vessel occlusion.

In spite of considerable pathophysiologic differences between cerebrovascular disease and ischaemic heart disease, the risk factors are largely the same for the two diseases.

3.1 Prevention in clinical practice

The ‘population strategy’ aims at shifting the distribution of risk factors towards more favourable levels through actions directed towards changing the environment and lifestyle of individuals, without the need to medically examine them, in order to reduce the incidence of disease by preventing or delaying the occurrence of acute cardiovascular events and the progression of chronic disease [76].

The high-risk primary prevention strategy deals with healthy persons with high absolute risks of future disease, and the secondary prevention strategy deals with patients with established cardiovascular organ damage or disease. The strategies aim to diminish the total cardiovascular risk of individuals belonging to the upper part of the risk distribution. The difference between the groups will be reflected in the priorities for intervention and the intensity of treatment. Since risk is a continuum, with many asymptomatic high risk people having investigative evidence of atherosclerosis, the terms ‘primary’ and ‘secondary’ prevention, while convenient, are artificial.

When dealing with prevention strategies based on individual interventions, it is necessary to bear in mind two facts. First, that to prevent one single cardiovascular event, it will be necessary to intervene upon many patients with no apparent benefit to them (prevention paradox). Second, that the number of patients needed to treat to prevent one case will vary in different populations or population subgroups (for example in women) depending on their underlying prevalences and distribution of risk factors and their incidence rate of disease. This problem is partly dealt with by using the estimation of absolute risk to trigger intervention, which helps to avoid over-treatment of low risk young women and under-treatment of high risk middle-aged men.

Coronary heart disease secondary prevention programs have proven to be effective in improving processes of care, readmissions to hospital, functional status and overall mortality, especially if they incorporate exercise programs. However, the effect sizes are quite modest and their cost effectiveness on a large scale remains uncertain [77, 78].

The term ‘screening’ refers to organized public health action targeted to examine the whole of the at risk or vulnerable population to diagnose earlier stages of disease in order to treat them under the assumption that earlier treatment will be more beneficial than treatment at later stages of disease development. It has been known for a long time that mass screening to prevent disease or its consequences has to fulfil certain conditions to be effective [79]. Screening has to be distinguished and not confused with opportunistic or systematic detection at a regular medical encounter. The term screening should not be used superficially. There is no evidence that mass screening for detection of early stages of CHD or stroke is a cost-effective way to prevent disease [80]. A different issue is screening for biological factors or for lifestyles associated with future disease. This type of screening should focus on those at high risk. But to identify those at high risk in the first instance, it is necessary to examine everyone in the population on a periodic basis, something unfeasible for any health system. Moreover, success of this type of action depends, among other things, on accurate and adequate systems of measurement, often lacking in real care settings, time for advice or treatment as appropriate, continuity of care and patients’ access to treatments, regardless of socioeconomic status. As many adults visit their GPs generally once a year, these visits can be used to identify those at high risk-in what is called opportunistic detection or case finding. This activity has implications for the doctor – patient relationship and it may imply ethical problems. In some cases, the patient will ask for risk assessment and counselling. In other cases, the doctor will identify risk factors and propose or offer further investigation and treatment. If the possibility of high cardiovascular risk is obvious, the doctor should offer counselling on lifestyle change and help but at the same time respect the patient's own values and choices. Advice should be offered but not imposed.

3.2 Policy issues

Aspects of health, economics and political action were considered in the Osaka Declaration arising from the Fourth International Heart Health Conference [81]. These may be summarized as:

In recent times, in the spirit of the Osaka Declaration, the ESC has joined with the European Heart Network to engage with the European Union to promote a coordinated European approach to the prevention of cardiovascular diseases. A conference of the relevant bodies was facilitated by the Irish Ministry of Health in February 2004. This informed the conclusions of the EU Council on Employment, Social Policy, Health and Consumer Affairs in June 2004 and an EU Heart Health Conference that resulted in the Luxembourg Declaration of 29 June 2005. This declaration defined the characteristics that are associated with cardiovascular health as:

Importantly, factors that are essential for implementation were defined, and the need for continued European Commission, Parliament and Council activity to promote a heart healthy Europe was stressed. A European Charter on Heart Health has been prepared. The text of the Luxembourg Declaration is available by searching in http://www.google.com for ‘Luxembourg declaration heart health charter’ [82].

The European Heart Health Charter is now available through the ESC website, escardio.org. It represents the combined efforts of the European Society of Cardiology, European Union and European Heart Network in close collaboration with the World Health Organization Regional Office for Europe. The charter acknowledges that cardiovascular disease is the number one cause of death among men and women in Europe. It affirms the characteristics associated with cardiovascular health defined in the Luxembourg Declaration above. It also categorizes the major risk factors. The signatories to the charter agreed to:

Further articles deal with education, health promotion, the establishment of national strategies for health promotion, and the promotion and adoption of the European Guidelines on Cardiovascular Disease Prevention. It is further agreed to prioritize research on the effectiveness of policy and prevention interventions, to assess the current status of cardiovascular health including risk factor prevalences to measure progress made at population and individual level to achieve the characteristics associated with cardiovascular health.

Fourteen major European bodies associated with cardiovascular health are signatories to the charter.

4.1 What is ‘evidence'?

The evaluation of interventions and diagnostic methods can make use of a wide range of sources of evidence: experience, retrospective case review, case reports, case series, historic and geographic comparisons, drug (and postmarketing) surveillance studies, pharmacoepidemiologic databases, cross-sectional studies, case-control studies, cohort studies, randomized controlled trials, and systematic reviews of trials and of observational studies. It is clear that different questions require different scientific methods, and that reliance on one source of evidence to the exclusion of others is likely to be misleading. This is particularly true in cardiovascular disease prevention. Lifestyle measures such as smoking cessation, exercise and healthy eating are less amenable to double blind randomized control trials than are drug treatments and too strict an adherence to the primacy of the randomized control trial may result in guidelines that promote excessive usage of drugs.

OPEN IN VIEWER

Table 2

Desirable attributes of clinical guidelines

Attribute of guideline	How to test the attribute
Validity	Are the health benefits/costs predicted achieved in practice?
Reproducibility	Does the same evidence and method result in the same recommendation?
Reliability	Do the guidelines result in the same interpretation in similar clinical circumstances?
Representative development	Were key groups affected by the guidelines involved in its development?
Clinical applicability	Can evidence be used to define the patients involved or does this require judgement?
Clinical flexibility	Are exceptions permitted and are patient preferences considered?
Clarity	Are the guidelines unambiguous and user friendly?
Meticulous documentation	Who does this, what assumptions are made, what evidence has been collected, and what methods have been used in compiling the guidelines?
Scheduled review	When and how will guidelines be reviewed and updated if necessary?

Traditionally, hierarchies of evidence have been promoted as a means of prioritizing recommendations, and these generally put systematic reviews at the top of the hierarchy and case reports at the bottom, on the grounds that the potential for bias gets progressively smaller as the hierarchy is ascended. But this approach may be misleading as the quality of the evidence ultimately depends on the question to be answered. Quality implies being fit for purpose – for example, the best evidence to identify rare hazards of treatment is not a systematic review or a randomized controlled trial. Case reports may provide the first hint that a treatment is hazardous but require confirmation in large prospective surveillance studies. Diagnostic tests are seldom evaluated in randomized controlled trials as the question is not usually ‘Does use of the test improve outcomes?', but more commonly, what is the sensitivity and specificity of this test compared with the best method.

The National Institute for Health and Clinical Excellence (NICE) in the United Kingdom and other guideline developing bodies have adopted the ‘hierarchy of evidence approach’ in grading evidence of effectiveness, resulting in grades of recommendation derived from the level of evidence. This approach is difficult to use in practice as it is easy to conflate the study design with the quality of evidence, such that a systematic review may be always viewed as ‘stronger’ evidence than a randomized trial or an observational study. Application of the hierarchy of evidence method requires explicit judgements to be made about the quality of the evidence (e.g. completeness, potential for bias, adequacy of outcomes assessed, etc.). Critically, both the benefits and the hazards of interventions need to be taken into account in producing clinical guidance. This is best done if a ‘balance sheet’ of benefits and harms is prepared that allows expert opinion, policy makers and users of services to make informed judgements in grading the strength of a recommendation.

Linking the strength of evidence underpinning a recommendation (e.g. Grade A – mandatory to implement; Grade B – implement if resources permit; Grade C-limited evidence in support) is difficult to do explicitly as the process of formulating guidance, in distinction from that of preparing synthesis of the evidence, is less well documented and more informal. Importantly, confusion arises when recommendations appear to be made on the basis of the same evidence but come to different conclusions which may reflect that different questions have been posed, additional information such as cost-effectiveness estimates have been included, or that bias has arisen in the process of conducting the review. Methods that might be used to make the process more explicit include consensus approaches or voting, but these are seldom used in practice.

4.2 The problems of evidence and guidance

In using evidence to produce guidance or recommendations for clinical or public health practice, it is important to distinguish between the quality of the evidence (is it robust, little likelihood of bias, generalizable etc.) and the strength of a recommendation underpinned by the evidence. Not all high quality evidence merits a strong recommendation.

In England and Wales, NICE currently uses a modified hierarchy of evidence approach based on the SIGN (Scottish Intercollegiate Guidelines Network) [88] grading system for intervention studies and no longer grades recommendations [89]. This change was spurred by the potential for misinterpretation of letter grades at local levels, leading to unintended problems in implementation of programmes of care. For example, in implementing a guideline, some recommendations that are crucial to the overall improvement of care may be underpinned by little or no strong evidence of benefit, and would consequently be graded as low level in terms of strength of evidence. Managers determined to implement only Grade A recommendations on grounds of resource constraints would find that the intended integrated implementation of a care pathway – for atrial fibrillation, for example – would not work properly.

A major problem in abandoning linking of grades of evidence with recommendations is that transparency is lost and it is not easy to work out what is based solely on expert opinion or lobbying by groups with special interests. Consequently, there is interest in developing a system that retains the desirable ability to maintain transparency but avoids perversity in implementation. NICE now uses a system of key recommendations which are derived from a prioritization process taking into account recommendations that: have high impact on patients’ outcomes, including mortality and morbidity; have a high impact on reducing variation; lead to a more efficient use of NHS resources; and, if implemented would mean patients reach critical points in the care pathway more quickly. To ensure that there is transparency in the link between evidence and recommendations, guidance now contains structured evidence to recommendation statements describing how judgements were made, a process that is believed to be more helpful and transparent than a letter grade.

The World Health Organization, with input from NICE and others, have established a working group attempting to develop such a system, called GRADE [90]. This system does not use letter codes but grades evidence in terms of how trustworthy it is. Briefly, the evidence is classified by the outcomes relevant to the question being answered and the evidence for specific outcomes is appraised in four domains: study design, study quality, consistency and directness. The latter attribute is defined as whether the evidence is directly applicable to the people of interest – in epidemiology this is termed ‘generalizability’ and NICE calls it ‘applicability'. Then on the basis of explicit scoring of these attributes, the evidence quality for the specific outcome is defined as ‘high’ (further research unlikely to change confidence in the estimate of effect), ‘moderate’ (further evidence is likely to have an important impact on confidence in the estimate of effect), ‘low’ (further evidence is very likely to have an important impact on confidence in the estimate of effect) and ‘very low’ (any estimate of effect is very uncertain).

In making recommendations, the GRADE system asks the question does the intervention do more harm than good? Here it is suggested that the answers to this question can be categorized as: net benefits, trade-offs, uncertain trade-offs, and no net benefits. Finally, recommendations should be presented as do it/don't do it or probably do it/probably don't do it [90]. Clearly, GRADE cannot tackle the problems of implementation discussed above. The advantage of the GRADE system is that it makes the judgement of whether the beneficial effects of an intervention outweigh its unwanted effects which is considered to provide an explicit and transparent method of prioritizing interventions for implementation. It is likely that as expertise with GRADE grows and its operational practices mature and stabilize, it will become a more widely used tool for the generation of clinical guidance [91]. Certainly at least one American College is taking it up as the means of producing its recommendations in respiratory diseases [92].

In some circumstances, these implementation problems are not an issue as a guideline development group may not have responsibility for implementation, is not intending to produce comprehensive or integrated guidance on a specific topic, or has a role only in synthesizing evidence upon which other groups will make the recommendations appropriate to local circumstances. The European task force represents a guideline group of this nature. Its role is to synthesize published evidence on each area of management of cardiovascular disease which can then be used by national groups to develop locally appropriate guidance. It is clearly important that both the quality of the evidence and the strength of a recommendation that flows from it are assessed and graded explicitly. This approach ensures that strong recommendations in the face of weak evidence are clear and the judgements made in coming to such conclusions may then be more critically appraised by implementation groups.

We have attempted to ensure that the most appropriate evidence is used to underpin recommendations. For population prevention programmes observational epidemiological findings are an important first step in considering causality. Behaviours such as smoking cessation and exercise are less amenable to randomized control trials than drug treatments. Clearly, systematic reviews (http://www.cochrane.org) of observational studies are preferable to citation of single observational studies. For example, individual studies of the relationship between homocysteine and cardiovascular disease have demonstrated variable associations [93]. Pooling data can provide greater understanding of sources of heterogeneity introduced either by study design (e.g. case-control versus cohort) or by the nature of the participants and will provide a more precise estimate of effect. However, it is important to be aware that this increased precision may be spurious if the control for confounding and other biases is weak in the index studies [94].

A further and growing concern in epidemiology is that with some associations causation has been wrongly attributed. This appears to be the case for antioxidant vitamins where observational studies suggested a reasonable protective effect, but randomized controlled trials have shown that the interventions may even be harmful [95, 96]. Similar concerns have now become apparent with hormone replacement therapy that was thought to confer benefits, but an early systematic review [97] showing adverse cardiovascular effects was ignored until recent randomized controlled trials of hormone replacement therapy confirmed this adverse effect [98].

A further concern for us is the nature of available evidence. Much of the evidence concerns drug treatments rather than lifestyle interventions or health system improvements. Since robust evidence from systematic reviews of randomized controlled trials exists for benefits of statins on cardiovascular disease outcomes [99], the use of such drugs may receive more emphasis than, for example, smoking cessation.

In examining the effects of interventions, we have given prominence to Cochrane systematic reviews where they exist, as these are conducted to a rigorous standard and are updated periodically. We have used other systematic reviews where these exist and have only cited individual trials where they make particular points of interest, or are sufficiently large to provide a clear answer to a clinical question. Where we feel the evidence is scant we have stated this.

When examining effect sizes we have not used numbers needed to treat as these have quite marked problems [100], particularly in preventive cardiology where baseline rates of cardiovascular disease vary markedly throughout Europe. Consequently, a number needed to treat would be needed for countries with low, medium and high risk. Moreover, numbers needed to treat for different age groups and for men and women would be required. Relative risk reductions of treatment are applicable to all European populations, age groups and men and women as, in general, most treatments have the same relative benefits at different levels of risk.

In this report we have attempted to follow an evidence-based approach. We have defined the following questions:

We have systematically and critically reviewed the relevant literature to answer each question posed. Efforts have been made to implement the guidelines through the various participating societies. Previous guidelines have been evaluated by means of EUROASPIRE I and II [101]. In the future it is likely that our guidelines will be formulated using the GRADE system [90] or some variant of it.

The issues raised in this chapter raise certain difficulties with regard to the current ESC hierarchical grading system. As noted in this chapter and the preamble, the present system is likely to favour drug treatments over major lifestyle measures because the latter are less amenable to double blind randomized control trials. For this reason, after prolonged debate, the Task Force has not included tables of the grades that it prepared. However, it is anticipated that this issue will require further debate.

5.1 Introduction

At the outset, it is stressed that these guidelines are just that, and not didactic rules. They should be interpreted in the light of the clinician's own knowledge and judgement, the patient's view, and in the light of local conditions and practicalities and as new knowledge becomes available. Indeed the development of national guidelines is strongly encouraged with objectives, priorities and implementation strategies that are adapted to suit local conditions, both medical and economic.

The PRIORITIES suggested are to assist the physician in dealing with individual people and patients. As such, they acknowledge that individuals at the highest levels of risk gain most from risk factor management. As noted elsewhere, although such individuals gain most, most deaths in a community come from those at lower levels of risk, simply because they are more numerous compared to high risk individuals who, paradoxically, develop fewer events in absolute terms – the Rose Paradox [102] (Fig. 1). Thus a strategy for individuals at high risk must be complemented by public health measures to reduce, as far as is practicable, population levels of cardiovascular risk factors and to encourage a healthy lifestyle.

The encouragement of TOTAL RISK ESTIMATION as a crucial tool to guide patient management has been a cornerstone of the Guidelines since the first (1994) edition [1]. This is because clinicians treat whole people (and not individual risk factors), whose cardiovascular risk usually reflects the combined effects of several risk factors that may interact, sometimes multiplicatively.

Although clinicians often ask for thresholds to trigger intervention, this is problematic since risk is a continuum and there is no exact point where, for example, a drug is automatically indicated. This issue is dealt with in more detail, as is the issue of how to advise younger persons at low absolute but high relative risk, and the fact that all elderly people will eventually be at high risk of death and may be over-exposed to drug treatments.

The overall OBJECTIVES of cardiovascular prevention are to reduce mortality and morbidity in those at high absolute risk and to assist those at low absolute risk to maintain this state, through healthy lifestyle. Here, the risk charts are helpful – if blood pressure is hard to fully control, for example, total risk can still be reduced by stopping smoking or perhaps reducing cholesterol levels further. Although thresholds for total cardiovascular risk included in this guideline are arbitrary, targets for individual risk factors are even more problematic in that they will always be open to debate, are not always achievable and, notably, also because they seem to promote a uni-risk factor approach to prevention.

OPEN IN VIEWER

Figure. 1

The expected number of CVD deaths at increasing levels of predicted risk (expressed in % over 10 years). Illustration of the fact that most events occur in low risk people simply because they are more numerous compared to high risk individuals who, paradoxically, develop fewer events in absolute terms.

Yet clinicians ask for guidance so an attempt to define desirable levels of individual risk factors has been made in the context of more specific objectives.

5.2 Priorities

Individuals at highest risk gain most from preventive efforts, and this guides the following priorities:

In general, a middle-aged person with a 10-year risk of CVD death of 5% or more, is regarded as at high risk. Examination of the FINRISK MONICA data (which contribute substantially to the SCORE high risk population charts) suggests that the equivalent combined fatal + nonfatal CVD risk is about 10% – more in younger men and less in women and the elderly. The Dutch guidelines on CVD risk management present an adaptation of the SCORE risk function which estimates the risk of fatal and nonfatal CVD, based on Markov modelling. This shows that with increasing risk of fatal CVD the relative increase in the risk of fatal and nonfatal CVD is smaller. At 5% risk of fatal CVD the risk of fatal and nonfatal CVD events was 10% in this model [103]. The likelihood of requiring medication in addition increases with increasing risk.

5.3 Total risk estimation

Cardiovascular risk in the context of these guidelines means the likelihood of a person developing an atherosclerotic cardiovascular event over a defined period of time.

'Total risk’ implies an estimate of risk that is made by considering the effect of the major factors age, gender, smoking, blood pressure and lipid levels. While the term has become widely used, it is acknowledged that it is not comprehensive in that the effects of other risk factors are not considered except as qualifying statements.

OPEN IN VIEWER

Table 3

Impact of combinations of risk factors on risk

Sex	Age (years)	Chol (mmol/l)	BP (mmHg)	Smoke	Risk %
F	60	8	120	No	2
F	60	7	140	Yes	5
M	60	6	160	No	8
M	60	5	180	Yes	21

The importance of total risk estimation before management decisions are made is illustrated in Table 3 and Fig. 2. The figure illustrates that the effect of lipid levels on risk is modest in women who are at otherwise low risk, and that the risk advantage of being female is lost by the combination of smoking and mild hypertension. Table 3 shows that a person with a cholesterol of 8 mmol/l can be at 10 times LOWER risk than someone with a cholesterol of 5 mmol/l if the latter is a male hypertensive smoker. Randomized controlled drug trials of single risk factors do not give sufficient data to fully address these issues. While audits such as EuroAspire [101] suggest inadequate risk factor management in very high risk patients, it is also likely that, in the context of low risk patients who have not had a vascular event, there is the potential for substantial overuse of drugs by inappropriate extrapolation of the results of trials conducted mostly on high risk men to low risk individuals. In general, women and old and young patients have been underrepresented in the classic drug trials that have informed guidelines to date.

For these considerations to have an impact on clinical practice, it is essential for the clinician to be able to assess risk rapidly and with sufficient accuracy to allow logical management decisions.

This realization led to the development of the risk chart used in the 1994 and 1998 Guidelines [1, 2]. This chart, developed from a concept pioneered by Anderson [104] used age, sex, smoking status, blood cholesterol and SBP to estimate the 10-year risk of a first fatal or nonfatal coronary heart disease event. There were several problems with this chart. First, it was derived from American data from the Framingham study and the applicability of the chart to all European populations was uncertain. Second, the data set used was fairly small. Third, the definitions of nonfatal CHD events differed from those used in many other studies making it difficult to validate the chart. Finally, estimation of the risk of other manifestations of atherosclerosis such as stroke or aneurysm of the abdominal aorta was not possible.

The 2003 Guidelines [3] used a new system for risk estimation called SCORE (Systematic COronary Risk Estimation) [105], based on data from 12 European cohort studies and includes 205 178 patients examined at baseline between 1970 and 1988 with 2.7 million years of follow-up and 7934 cardiovascular deaths. The SCORE risk function has been externally validated using different datasets [106].

OPEN IN VIEWER

Figure. 2

The relationship of total cholesterol/HDL-cholesterol ratio to 10-year fatal CVD events in men and women aged 60 years with and without risk factors, based on a risk function derived from the SCORE project.

Risk charts such as SCORE are intended to facilitate risk estimation in ostensibly healthy persons. Patients who have had a clinical event such as an acute coronary syndrome or stroke have already declared themselves to be at high risk of a further event and automatically qualify for intensive risk factor evaluation and management.

SCORE differs from earlier risk estimation systems in several important ways, and has been modified somewhat for the present guidelines:

In the 2003 guidelines [3], a 10-year risk of CVD death of 5% or more was arbitrarily considered high risk. Yet this implies a 95% chance of not dying from CVD within 10 years, less than impressive when counselling patients. The new nomenclature in this 2007 guideline is that everyone with a 10-year risk of CV death of 5% or more has an increased risk. Naturally the risk of total fatal and nonfatal events is higher, and clinicians naturally wish for this to be quantified. The biggest contributor to the high risk SCORE charts is FINRISK, which has data on nonfatal events defined according to the MONICA project [107]. Calculating total event rates from FIN-RISK suggests that, at the level (5%) at which risk management advice is likely to be intensified, total event risk is about 10%, more (15%) in younger men and somewhat less in women and in older persons.

As noted in the introduction, clinicians often ask for thresholds to trigger certain interventions, but this is problematic since risk is a continuum and there is no threshold at which, for example, a drug is automatically indicated. A particular problem relates to young people with high levels of risk factors – a low absolute risk may conceal a very high relative risk. In the 2003 Guidelines [3] it was suggested to extrapolate risk to age 60 to stress that a high absolute risk would occur if preventive action were not taken. It was not intended that such a young person should be necessarily treated as if they were 60, but a literal interpretation of this suggestion could lead to excessive drug treatment in younger persons. This part of the text has been rephrased, and a relative risk chart added to the absolute risk charts to illustrate that, particularly in younger persons, lifestyle changes can reduce risk substantially as well as reducing the increase in risk that will occur with ageing.

Another problem relates to old people. In some age categories the vast majority, especially of men, will have estimated CV death risks exceeding the 5 to 10% level, based on age (and gender) only, even when other CV risk factor levels are relatively low. This could lead to excessive usage of drugs in the elderly. This issue is dealt with later in this section.

As before, charts are presented for both total cholesterol and cholesterol: HDL-cholesterol ratio. They look remarkably similar. However, subsequent work on the SCORE data base, as yet unpublished, has shown that HDL-cholesterol can contribute substantially to risk prediction if entered as an independent variable.

Dealing with the impact of additional risk factors such as HDL-cholesterol, body weight, family history and newer risk markers is difficult within the constraint of a paper chart. The electronic, interactive version of SCORE, HeartScore (available through escardio.org)is not so constrained. It presently replicates SCORE in electronic format but will be used to accommodate the results of new SCORE analyses, such as those relating to HDL-cholesterol as these are checked and validated. It should be stressed, however, that although many risk factors other than the few included in the available risk functions have been identified [such as C-reactive protein (CRP) and homocysteine levels] their contribution to absolute CV risk estimations of individual patients (in addition to the older risk factors) is generally modest.

The impact of self-reported diabetes has been reexamined. While there is heterogeneity between cohorts, overall, the impact of diabetes on risk appears greater than in risk estimation systems based on the Framingham cohort, with relative risks of approximately five in women and three in men (unpublished data).

Some of the advantages of using the risk charts may be summarized:

ADVANTAGES IN USING THE RISK CHART

The SCORE risk charts are shown in Figs 3–7, including a chart of relative risks. Instructions on their use and qualifiers follow.

This relative risk chart is useful in explaining to a younger person that, even if his/her absolute risk is low, it may still be 10-12 times higher than that of a person of a similar age with low risk factors.

5.3.1 How to use the risk estimation charts

•

The low risk charts should be recommended for use in Belgium, France, Greece, Italy, Luxembourg, Spain, Switzerland and Portugal and also in countries which have recently experienced a substantial lowering of the CV mortality rates. The high risk charts should be recommended in all other countries of Europe. Note that several countries have undertaken National recalibrations to allow for time trends in mortality and risk factor distributions. Such charts are likely to better represent current risk levels.

•

To estimate a person's 10-year risk of CVD death, find the table for their gender, smoking status and age. Within the table find the cell nearest to the person's blood pressure and total cholesterol or cholesterol: HDL-cholesterol ratio. Risk estimates will need to be adjusted upwards as the person approaches the next age category.

•

Low risk persons should be offered advice to maintain their low risk status. While no threshold is universally applicable, the intensity of advice should increase with increasing risk. In general, those with a risk of CVD death of 5% or more qualify for intensive advice, and may benefit from drug treatment. At risk levels over 10% drug treatment is more frequently required. In persons older than 60 these thresholds should be interpreted more leniently, because their age-specific risk is normally around these levels, even when other CV risk factor levels are ‘normal'. In particular, uncritical initiation of drug treatments of all elderly with risks beyond 10% threshold should be discouraged.

•

Relative risks may be unexpectedly high in young persons, even if absolute risk levels are low.

•

The charts may be used to give some indication of the effects of reducing risk factors, given that there will be a time lag before risk reduces and that the results of randomized controlled trials in general give better estimates of benefits. Those who stop smoking in general halve their risk.

OPEN IN VIEWER

Figure. 3

SCORE chart: 10-year risk of fatal CVD in populations at high CVD risk based on the following risk factors: age, gender, smoking, systolic blood pressure, total cholesterol.

OPEN IN VIEWER

Figure. 4

SCORE chart: 10-year risk of fatal CVD in populations at low CVD risk based on the following risk factors: age, gender, smoking, systolic blood pressure, total cholesterol.

OPEN IN VIEWER

Figure. 5

SCORE chart: 10-year risk of fatal CVD in populations at high CVD risk based on the following risk factors: age, gender, smoking, systolic blood pressure, total cholesterol/HDL-cholesterol ratio.

OPEN IN VIEWER

Figure. 6

SCORE chart: 10-year risk of fatal CVD in populations at low CVD risk based on the following risk factors: age, gender, smoking, systolic blood pressure, total cholesterol/HDL-cholesterol ratio.

OPEN IN VIEWER

Figure. 7

Relative risk chart.

The relative risk chart may be helpful in identifying and counselling such persons.

5.4 Objectives of CVD prevention

5.5 Conclusions

The PRIORITIES defined in this section are for clinical use and reflect the fact that those at highest risk of a CVD event gain most from preventive measures. This approach should complement public actions to reduce community risk factor levels and promote a healthy lifestyle.

Estimation of TOTAL RISK remains a crucial part of the present guidelines. The SCORE system has been updated with an estimate of total CVD risk as well as risk of CVD death. New information on diabetes is included. Information on relative as well as absolute risk is added to facilitate the counselling of younger persons whose low absolute risk may conceal a substantial and modifiable age-related risk.

The difficulty in imposing arbitrary thresholds or targets upon a continuous variable such as risk is acknowledged. Nevertheless, specific OBJECTIVES are defined in terms of desirable levels of individual risk factors. This must be seen as an aid to clinicians in planning risk management strategies with their patients. The primacy of managing total risk rather than focusing on individual risk factors is stressed.

Priorities, risk estimation and the definition of objectives reflect an attempt to make complex issues simple and accessible. Their very simplicity makes them vulnerable to criticism. Above all they must be interpreted in the light of the physician's detailed knowledge of their patient and in the light of local guidance and conditions.

A summary of the recommendations is given in the flow chart below:

OPEN IN VIEWER

Figure

When do I assess total CVD risk?

6.1 Strategies for promoting behavioural change

Physicians and other health professionals in the primary and out-patient care setting are in a unique position to contribute significantly to the improved prevention and management of CVD. Physicians are generally perceived by the general public as the most reliable and credible source of information on health and advice. Patients usually want to receive as much information as possible from physicians, and often prefer to receive assistance from them in order to change behaviours such as smoking, nutrition and diet, and physical activity, rather than to attend special programmes elsewhere [108].

6.2 The physician/caregiver-patient interaction as a means towards behavioural change

A friendly and positive physician-patient interaction is a powerful tool to enhance patients’ coping with stress and illness and adherence to recommended lifestyle change and medication. Social support is known to exert positive influences on illness behaviour, on coping and on adherence; conversely, a nonsupportive environment can lead to a cycle of misunderstandings and destructive emotions (e.g. rage and anger), which in turn result in attention shifts, motivational problems and ignorance of the threats of chronic illness and the need to change lifestyle.

Social support provided by caregivers, including physicians, may be of primary importance to help patients maintain healthy habits and follow medical advice. Decision-making should be shared between physician/ caregiver and patient (also including the patient's spouse and family) to the largest extent possible, thus assuring patients’ and families’ active involvement in lifestyle change and medication adherence.

The physician's use of some principles of effective communication will facilitate successful treatment and prevention of CVD.

6.3 Interventions to improve adherence

Preventive interventions can only exert their effects if patients actually adhere to them. Recent meta-analytic data show that adherence with beneficial medications is associated with improved survival. However, adherence with placebo also improves survival [111] indicating that it may reflect generally better health behaviour. Measures should therefore be taken to improve adherence and health behaviour in general. Motivational interviewing and counselling based on the individual situation of the patient and his or her readiness to adopt behaviour changes increases the likelihood of these changes taking place and shared decision making can facilitate the maintenance of measures agreed upon.

6.4 Evidence for psychosocial risk factor modification programs

A number of psychosocial intervention strategies have been demonstrated to have positive effects on risk factors (see specific chapters on nutrition, smoking, physical activity and psychosocial factors) – but the specific content and approaches taken by these interventions vary. Even if they intend to target only one behavioural risk factor, such as low social support, group-based behavioural interventions often contain elements, which affect multiple risk factors [112, 113]. For example, cardiovascular patients in a smoking cessation program may also benefit from social support or behavioural advice from their peers. Interventions adding psychosocial and psycho-educational components to standard cardiologic care can significantly improve quality of life and diminish cardiovascular risk factors. Nevertheless, data are inconclusive for effects of psychosocial interventions on disease progression and survival [114]. Such programs that achieve their proximal goals, that is reduction of behaviour-dependent risk factors, also seem to have the potential to prevent the progression of clinical CHD. There is some evidence that intervention programs should be individualized based on individual risk constellations and include gender-specific aspects [115].

6.5 Multimodal interventions

The physician/caregiver should recognize the social, emotional and cognitive problems associated with illness and lifestyle change and help to develop strategies to solve them. Continuing education curricula to increase physicians’ behavioural and psychosocial skills have been developed in some countries [116–119] and may help to increase the effectiveness of counselling and behavioural advice. It should, however, be recognized that the physician is not the only professional person involved. The expertise of psychiatry, psychology, psychosomatic medicine, nutrition and behavioural sciences in a broader sense is also needed. In particular, therapeutic and preventive interdisciplinary team work should be attempted, combining the appropriate knowledge and skills to optimize the preventive efforts.

Multimodal, behavioural interventions are especially recommended for patients with clinically manifest CVD and for individuals at very high risk. These interventions should integrate education on healthy lifestyle and medical resources, exercise training, relaxation training, and smoking cessation programmes for resistant smokers. Whenever needed, additional individual or group counselling should be performed. Such counselling can enhance coping with illness, improve adherence with prescribed medication and efforts to change behaviour, and facilitate adequate utilization of medical resources, in particular to minimize delay in seeking help in case of central chest pain or other serious symptoms. Psychosocial risk factors (stress, social isolation and negative emotions) that may act as barriers against behaviour change will also be addressed in individual or group counselling sessions, according to specific needs of the participants.

Multimodal interventions need special education of the staff. There is evidence that more extensive/longer interventions may lead to better long-term results with respect to behaviour change and somatic outcome [120]. Patients of low socio-economic status, of older age, or female gender may need tailored programmes, in order to meet their specific needs regarding information and emotional support.

7.1 Smoking as a risk factor

There is overwhelming evidence for an adverse effect of smoking on health [121]. In long-term smokers, smoking is responsible for 50% of all avoidable deaths and one half of these are due to cardiovascular disease [122, 123]. Tobacco smoking causes heart attacks at any age, not only in the elderly; in the MONICA project more than half of all nonfatal myocardial infarctions in young people aged 35-39 years could be attributed to smoking [124]. This adverse effect of smoking is related to the amount of tobacco smoked daily and to the duration of smoking [125, 126]. The risk of future cardiovascular disease is particularly high if smoking starts before the age of 15 years [127]. Initially, smoking of tobacco was mainly taken up by men but within the past decades male and female smoking habits have become more similar in Europe. In recent prospective studies the mortality from vascular disease has been found to be proportionately higher in female smokers than in male smokers [128, 129]; this difference remains significant after adjustment for major CV risk factors. A higher risk of lung cancer in smoking women as compared to men has also been described in several studies. This could be related to differences in nicotine metabolism and smoking-related behaviours. Women metabolize nicotine faster than men, especially women taking oral contraceptives [130]. Women under oral contraceptive treatment should specifically avoid smoking as cigarettes have a synergistic effect on the risk of both ischaemic heart disease and cerebral thromboembolism [131].

There have also been some questions on the relative strength of the association between smoking and CVD within ethnic subgroups; there have been some suggestions that Asians may have less excess risk from smoking than Caucasians. However, recent findings have shown that Asians and the Caucasian Australians and New Zealanders have similar increased proportionial CV risk from smoking cigarettes and similar relative risk reduction from quitting [128].

The impact of smoking on atherosclerosis progression is also greater for patients with diabetes and hypertension [132]. Smoking is more common among those who have received little education and widening education-related inequalities in smoking cessation rates have been observed in many European countries in recent years [133].

Within Europe, the impact of smoking on the absolute risk of coronary heart disease has been found to be smaller in Mediterranean populations than in Northern European populations [134]. Dietary factors probably explain this difference in the effect of smoking.

Passive smoking has now been shown to increase the risk of coronary heart disease and other smoking-related diseases [132, 135–137] and the effects of second-hand smoke on the cardiovascular system may even be greater than what was expected [138]; some of these effects appear rapidly and can precipitate acute manifestations of CVD in patients [139]. While cigarette smoking is certainly the most deleterious of the tobacco habits, other uses of tobacco, such as different forms of smokeless tobacco, are also related to health problems [140].

Although the exact mechanisms by which tobacco smoking increases the risk of atherosclerotic disease are not yet fully understood, smoking enhances both the development of atherosclerosis, the occurrence of superimposed thrombotic phenomena and leukocyte activation [141, 142]. Some of these effects may be even more important, because stopping smoking leads to a quicker reduction in the risk of subsequent coronary heart disease events in patients with established coronary heart disease than in asymptomatic individuals; in patients with established coronary heart disease the risk falls within 2-3 years to the level of those coronary heart disease patients who never smoked [126], whereas in asymptomatic individuals up to 10 years are needed to reach the risk level of those who never smoked [127, 143].

In a meta-analysis of cohort studies on the effect of smoking cessation on mortality after a myocardial infarction, all studies showed a mortality benefit of 0.54 with a combined odds ratio in those who quit (95%CI 0.46-0.62). The mortality benefit was consistent regardless of sex, duration of follow-up, study site and time period [144]. Therefore stopping smoking after a myocardial infarction is potentially the most effective of all preventive measures. Sufficient efforts should be devoted to this end.

7.2 Smoking assessment

The assessment of the smoking status of people should be done at every opportunity. Smoking history should include the following questions: Is the person a current smoker? If yes, number of cigarettes or grams of tobacco (cigars, pipes) smoked daily; life-time duration of smoking; earlier attempts to stop. If the person has stopped smoking, for how many years or months has he/she stopped?

In addition, in some individuals it may be of interest to know their degree of addiction and their state of mind with regard to change. The Fagerström test for nicotine dependence can be used for the latter [145].

7.3 Prevention and management of smoking

Stopping smoking should be encouraged in all smokers. There is no age limit to profit from the benefits of smoking cessation. Some of the advantages are almost immediate, others take more time. Figure 8 outlines the World Health Organization's recommended approach to the individual patient.

The benefits of smoking cessation have been extensively reported [121, 146]. Tobacco cessation strategies should involve behavioural, pharmacological and community-oriented interventions.

At the level of the individual patient a combined behavioural and pharmacological intervention may be the most appropriate, yet the public health impact may be greater by community-oriented interventions.

Quitting smoking is a complex and difficult process, because this habit is strongly addictive both pharmacologically and psychologically. Despite this, many people who succeed in quitting manage to do this without any special programme or treatment. The chances for becoming an ex-smoker at 6 or 12 months are, however, slim, somewhere around 1-3%. These results can be improved by professional assistance.

OPEN IN VIEWER

Figure. 8

WHO Smoking Cessation Algorithm.

The physician's firm, explicit but sympathetic advice that a patient with coronary heart disease or other atherosclerotic disease should stop smoking is the most important factor in getting the smoking cessation process started. The momentum for smoking cessation is particularly strong at the time of diagnosing athero-thrombotic cardiovascular disease and in connection with an invasive treatment, such as coronary artery bypass grafting, percutaneous transluminal coronary angioplasty or vascular surgery. The physician's advice is equally important in helping healthy high-risk individuals to attempt quitting smoking. The physician's explicit advice to quit smoking completely and ascertainment that the person is willing to try to do it, or at least to start the process of contemplating such an action, are the decisive first steps. Brief reiteration of the cardiovascular and other health hazards of smoking, providing appropriate literature, and agreeing on a specific plan with a follow-up arrangement are the essential features of the brief advice version of smoking cessation in clinical practice. This may involve the assessment of the degree of addiction and their stage of change. Pooled data from 17 trials of brief advice versus no advice (or usual care) revealed a small but significant increase in the odds of quitting (odds ratio 1.74, 95% CI 1.48-2.05). This equates to an absolute difference in the cessation rate of about 2.5% [147]. Both individual and group behavioural interventions are effective in helping smokers to quit as well [148–150]. A recent randomized trial in patients with acute cardiovascular disease found that a minimum of 12 weeks of behaviour modification counselling and individualized pharmacotherapy were superior to simple counselling plus printed educational materials not only in terms of 2-year smoking cessation; patients in the intervention group also had lower rates of rehospitalization and mortality than patients receiving simple advice [151].

Readers are referred to specific recommendations describing the principles of brief advice and other interventions for smoking cessation in clinical practice [145, 152]. At hospital-based clinics and primary healthcare practices nurses are an important resource in individual counselling on smoking cessation. Physicians and nurses need to set an example for their patients by not smoking themselves. Primary pipe or cigar smokers may be at somewhat lower cardiovascular risk than cigarette smokers, mainly because many of them tend to be noninhalers. It is advisable to try to get patients with atherosclerotic disease and high-risk individuals to also stop these forms of smoking. If cigarette smokers shift to pipe or cigar smoking, they usually continue to inhale and therefore this shift should be discouraged.

Nicotine chewing gum and the transdermal nicotine patch have been widely used in helping quitters to go through the difficult initial weeks or months of smoking cessation. A Cochrane review on the effectiveness of different forms of nicotine replacement therapy (NRT) revealed that the quit rates increase approximately 1.5 to 2-fold with the use of nicotine replacement [153]. Initial success is often followed by a relapse, but cessation rates of 10% or more for 1 year or longer have been achieved following nicotine replacement therapy. The use of nicotine patches has been successfully tested in patients who have coronary heart disease without any adverse effects [145].

The effectiveness of antidepressant medication in aiding long-term smoking cessation has been reviewed [154]. From this it seems that bupropion and nortriptyline can aid smoking cessation. In one trial the combination of bupropion and nicotine patch produced slightly higher quit rates than the patch alone.

Another new pharmacological agent that may be of help in smoking cessation is varenicline, a nicotine acetylcholine receptor agonist. Among long-term smokers treatment with varenicline was associated with a smoking cessation rate of 23% at one year as compared to 15 and 10.3% in the groups treated with bupoprion and placebo respectively [155, 156]. Reports that it may be more effective than buproprion or placebo need confirmation.

Support by the spouse and family is very important in smoking cessation. Involvement of the family in the smoking cessation process and getting other smoking family members to quit smoking together with the patient is of great help. There is no consistent evidence that acupuncture, acupressure, laser therapy or electrostimulation are effective for smoking cessation [157].

In many European countries a favourable development has occurred with the creation of ‘smoke-free’ environments, including restrictions of smoking at work sites, in public transport vehicles, restaurants, etc. These changes provide an improved atmosphere for smoking cessation attempts by individuals. The steps for effecting smoking cessation recommended by the WHO are detailed in Fig. 8 [158].

8.1 Scientific background

The role of nutrition in the aetiology and prevention of atherosclerosis and cardiovascular disease has been extensively reviewed [159]. Worldwide, there are strong, consistent and graded relationships between saturated fat intake, blood cholesterol levels and the mass occurrence of cardiovascular disease. The relationships are accepted as causal.

8.1.3 n-3 Fatty acids

α-Linolenic acid (ALA) is the precursor in the n-3 group. It is made up of 18 carbon atoms and three double bonds. The main food sources are certain vegetable oils: soybean, sunflower and linen oils. ALA is an essential fatty acid. In prospective epidemiological studies, a high intake of ALA is associated with a reduction in fatal cardiovascular events [169]. Data from randomized clinical trials on the effects of ALA on CVD events are limited and of poor quality [170].

Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are two significant representatives of the n-3 group [171]. These fatty acids are mainly derived from fish oils and fats. There is much evidence suggesting that consumption of EPA and DHA are beneficial for triglycerides, blood pressure, haemostatic balance and heart rhythm [172, 173]. Prospective epidemiological population studies show lower rates of fatal coronary occurrences and sudden death among people who regularly consume fish than among nonconsumers. A recent meta-analysis of cohort studies showed an 11% reduction of CHD mortality in individuals consuming fish one to three times per month, 15% reduction in those consuming fish once a week, 23% for intake of fish two to four times a week and 38% for five or more per week [174]. Similarly, a meta-analysis indicates that the intake of fish is inversely related to risk of stroke, particularly ischaemic stroke [175]. In general, the evidence for the benefits of fish oil is stronger in secondary than primary prevention [170].

Three secondary prevention trials have assessed the effect of fish or fish oil capsules on fatal coronary events. The GISSI study [176] showed a protective effect of EPA/DHA supplementation on fatal events in patients who had suffered a previous MI [RR for total mortality 0.59 (95% CI 0.36-0.97), n = 11 323]. The DART study [177] also showed an initial mortality benefit of dietary advice to increase oily fish intake. In the DART II trial [178], 3114 men with angina were randomized to diet advice to increase oily fish intake, advice to take fish oil supplements or no advice. In contrast to the other two large trials, DART II failed to demonstrate any beneficial effect. In fact, risk of cardiac death was higher in the intervention group (RR for cardiac death 1.26 (95% CI 1.00-1.58), the risk was mainly seen in those on the fish oil supplements. Many consider this study to be flawed, because it was not a randomized controlled trial, the trial was stopped and restarted and the supplement was not pure EPA/DHA and may have contained n-6 fatty acids, which are thought to interfere with the health benefits of n-3 fatty acids.

OPEN IN VIEWER

Table 4

Lipid lowering and high polyunsaturated fat intervention trials and coronary heart disease

				Changes
Trial	Subjects	n	Duration (y)	Change in cholesterol (%)	Difference CHD (%)a
Reduction in total fat
MRC	123	Ml patients	3	−5	4
DART	1015	Male Ml	2	− 3.5	−9
Women's Health Initiative Dietary Modification Trial	19541	Postmenopausal women	6-8	NS	NS
High Polyunsaturated fat diet
Finnish Mental Hospital	676	Men	6	−15	− 43∗∗
Minnesota Coronary Survey	9057	Men/women	4.5	−14∗∗	0
Los Angeles Veteran Study	424	Men	8	−13∗∗	−20 (−33∗∗ CVD events)
Oslo Diet-Heart Study	206	Male Ml	5	−14∗∗	−25∗∗
MRC	199	Male Ml	4	−16∗∗	− 12

^a% Difference in coronary events between treatment and control groups. ∗∗P >0.05. Adapted from Hu and Willett [161].

A recent Cochrane review [179] and meta-analysis [180], both by Hooper et al., concluded: ‘while it is unclear whether ω-3 fats alter cardiovascular end points, there is no evidence we should advise people to stop taking rich sources of ω-3 fats and further high quality trials are needed to confirm suggestions of a protective effect of n-3 fats on cardiovascular health.’ This meta-analysis was criticized mainly because the result was largely influenced by the inclusion of the DART II trial, which is considered to be flawed. On exclusion of this trial the pooled relative risk was similar to that seen previously [RR for all-cause mortality 0.87 (95% CI 0.73-1.03) with DART II, RR 0.83 (95% CI 0.75-0.91) without DART II]. The review was also criticized for considering trials with differing end points and differing patient populations together and for the inclusion of trials of plant and fish-derived n-3 fatty acids together. The above systematic review and the results of the DART II study should not be ignored. However, the many meta-analyses and RCTs showing a significant beneficial effect of ω-3 fatty acids on risk of CVD must also be considered. It is clear that further high quality RCTs are required.

Experimental, clinical and epidemiological data suggest that part of the benefit of n-3 fatty acids on fatal cardiovascular events could be due to antiarrhythmic or heart-rate regulating properties of these fatty acids [181]. In observational studies consumption of fish is associated with a lower incidence of atrial fibrillation in one [182] but not in another study [183]. A meta-analysis of randomized controlled studies with fish oil shows that fish oil reduces heart rate in humans [184]. The analyses on heart rate variability and baroreflex sensitivity has shown inconsistent results [185] andrandomizedclinicaltrials aimed at assessing the occurrence of premature ventricular complexes in patients with implantable defibrillators showed mostly no benefit from n-3 supplementation [185–188].

Recent studies have pointed out the possible harmful effects of methyl-mercury exposure on cardiovascular disease [189]. Because fishes accumulate methyl-mercury and fish consumption is a major source of mercury exposure in humans, several countries have posted fish consumption advisories to reduce exposure of sensitive subpopulation [190, 191].

8.2 Alcohol

Alcohol is not an essential nutrient. The pathophysiological consequences of its consumption depend on the conditions of use (chronic or acute), the quantities ingested, the pattern (with meals, only at weekends) as well as many individual factors (gender, age, genetic susceptibility etc.). Alcohol consumption is linked with an increase of haemorragic cerebrovascular accidents and, to a lesser extent, ischaemic stroke [201] which depends on the dose. On a population scale, the relationship between alcohol consumption and total mortality has a U or J shape. Reduced mortality related to moderate alcohol use results from reduced coronary mortality; an association that was attributed in large part to an effect on HDL-cholesterol, glucose intolerance and fibrinogen [202]. In contrast, observational studies and clinical trials have consistently demonstrated a direct, dose-dependent relationship between alcohol intake and blood pressure [203]. To date no randomized trial has proven that the voluntary intake of a moderate quantity of alcohol is beneficial in terms of cardiovascular morbidity and mortality. There is also no reliable proof showing any higher cardiovascular benefit of any drink, compared with another [204]. Furthermore, advocating the therapeutic cardioprotective use of alcohol raises serious concerns with regard to dependence and abuse with consequent toxicity.

8.3 Sodium, potassium and other electrolytes and blood pressure

There is abundant evidence from animal studies, clinical trials, epidemiological studies and meta-analyses that sodium intake increases blood pressure and therefore the risk of arterial hypertension, stroke, CHD and heart failure [205, 206]. In societies with low salt intake there is no age-related increase in blood pressure [207].

In a recent meta-analysis [208], a median reduction in urinary sodium excretion of 1.8 g/d lowered systolic and diastolic blood pressure by 2 and 1 mmHg in normoten-sive and 5 and 2.7 mmHg in hypertensive individuals.

Animal experiments, observational epidemiological studies and clinical trials support the evidence that high potassium intake is associated with reduced blood pressure. In a recent meta-regression, the average blood pressure reduction associated with a median −77 mmol/24 h increase in urinary sodium was 2.54 mmHg systolic and 1.96 mmHg diastolic [209]. The corresponding values for increased potassium intake of 44 mmol/24 h were −2.42 and −1.57 mmHg, respectively. Blood pressure response was larger in hypertensive than normotensive individuals, both for sodium and potassium. The effect of potassium on blood pressure depends on the concurrent consumption of salt. An increased intake of potassium has a greater lowering effect on a high salt intake background.

A recent Cochrane review on dietary salt reduction for the prevention of CVD [210] sounded a note of caution – intensive interventions, unsuited to primary care or population prevention programmes, resulted in minimal reductions in blood pressure during long-term trials. Further evaluations to assess effects on morbidity and mortality outcomes are needed for populations as a whole and for patients with elevated blood pressure. A low sodium diet may help to maintain a lower blood pressure after the withdrawal of antihypertensives.

8.4 Fruits and vegetables

Fruits and vegetables are significant sources of minerals, vitamins and fibres. Observational studies have shown favourable relationships between the consumption of the main antioxidant vitamins or of the plasma levels of such vitamins and cardiovascular diseases. These results have been confirmed in cohort studies which have shown negative correlations between the consumption of fruits or vegetables and the occurrence of stroke or coronary events [211]. In a meta-analysis of prospective observational studies each serving increment of fruit and vegetable intake was associated with a 4% reduction in risk of coronary events, respectively [212]. Similarly, in another meta-analysis, the consumption of each additional portion intake of fruit and vegetable per day was associated with a 5% reduction in the relative risk of stroke [213]. The results of these observational studies suggest that eating fruits and vegetables on a regular basis has beneficial effects on the cardiovascular system. However, in line with other lifestyle trials, a randomized controlled trial assessing the exact magnitude of this benefit would be difficult to conduct because of compliance and cross-over issues. This was the subject of a Cochrane review in 1999 [214].

The most important sources of potassium are fruits and vegetables. The effects of potassium on blood pressure are discussed above. The results of the DASH trials have demonstrated that the combined increase in the intake of fruit and vegetable and salt restriction was followed by greater reduction in blood pressure than the single effects of sodium restriction [205, 215–217]. Another trial used a brief negotiation method to encourage an increase the consumption of fruit and vegetables to at least five portions a day in patients aged 25-64 years without serious chronic illness. Systolic and diastolic blood pressure fell significantly by 4 and 1.5 mmHg, respectively, in the intervention group [218].

8.5 Whole grain intake

There is now evidence from large epidemiological studies which show a risk reduction of about 25-30% when whole grain consumption is at least three servings per day [196, 219–227]. There are no clinical trials to date. It is important to remember that carbohydrate intake contributes the major part of energy intake.

8.6 Vitamins supplements

8.7 Plant sterols

Plant sterols are naturally occurring constituents of plants that differ from cholesterol only in the structure of their side chain. The two most common constituents are sistosterol and campesterol and their saturated products sistostanol and campestanol, respectively. Previous findings indicate that plant sterol reduce cholesterol absorption by 30-50% [237]. This reduction is thought to result from the direct competition of plant sterols with cholesterol for incorporation into mixed micelles [238]. Because plant sterols are slightly more hydrophobic than cholesterol they are preferentially incorporated into the micelles in place of cholesterol. In order to be used in foods plant stanols need to be esterified. Consumption of 2 g/d of plant sterols has been shown to lower total and LDL-cholesterol in a variety of different populations [239]. In contrast, it has no impact on plasma triglycerides and HDL-cholesterol. In a dose-response study ranging from 0.8 to 3.2 g/d of stanol ester intake for 4 weeks, a significant decrease in LDL-cholesterol level was observed at doses 1.6 g/d or more. However, higher intakes at 3.2 g/d did not add further lowering compared to the 2.3 g/d dose. Further analyses reached the same conclusion and demonstrated an average of 10% reduction in LDL-cholesterol levels for an optimal intake of 2 g/d [240]. In general, the LDL-lowering effect of plant sterol esters is proportionately greater in individuals with high intakes of saturated fat and cholesterol than those with low intake. When associated with a high polyunsaturated fatty acid diet, LDL decreased by 19%. Few studies have analysed the efficacy of stanol esters in different dietary backgrounds. Most studies have used vehicles that were high in fat such as margarine. Recently, stanol esters were administered with low-fat yoghurt resulting in similar dose-to-dose LDL-lowering as stanol ester from margarine.

There are a number of concerns with the use of plant sterol and stanol esters. To date no randomized nutritional trial has yet proven that plant sterol or stanol esters confer a benefit in terms of cardiovascular mortality. Furthermore, plant sterol contents have been found to accumulate in atherosclerotic plaque [241]. Until long-term studies are performed to ensure safety in all individuals plant sterols should be regarded as optional, adductive therapy in adults with elevated LDL-cholesterol levels [240].

8.8 Policosanol

Policosanol, a natural mixture of aliphatic primar alcohols isolated from sugar cane, has been reported to lower total and LDL-cholesterol and increase HDL-cholesterol in healthy volunteers, in patients with hypercholesterolemia as well as hypercholesterolemic patients with type 2 diabetes [242]. The results of some studies with sugar cane policosanol even suggested similar efficacy to statins [243]. However, most of these studies were performed in a relatively small number of patients and in very few centres. There was just one study performed with rice policosanol which showed much less effect on total cholesterol and no effects on LDL-cholesterol, HDL-cholesterol and triglycerides [244].

8.9 Obesity and overweight: risk and management

8.9.2 Body weight and risk

It is now clear that fat, and in particular intra-abdominal visceral fat, is a metabolically active endocrine organ that is capable of synthesizing and releasing into the bloodstream an important variety of peptides and nonpeptide compounds that may play a role in cardiovascular homeostasis. Excess adipose tissue is associated with increased secretion of free fatty acids, hyperinsulinaemia, insulin resistance, hypertension and dyslipidaemia [246, 247]. This impacts on CVD risk factors and hence on risk. The mechanical effects of overweight impact on noncardiovascular cause of morbidity and mortality. The health effects of increasing body weight are summarized in Table 5.

Interestingly, the effects of multivariate adjustment on the association between lipid levels and risk and between body weight and risk are different. Raised blood cholesterol and reduced HDL-cholesterol level remain independently associated with risk after adjustment for other major risk factors, whereas the association between weight and risk tends to lose significance. This should not be interpreted as indicating that body weight is not important; rather, it may be critically important because it exerts its effect on risk by its adverse effects on many risk factors.

OPEN IN VIEWER

Table 5

Impact of increasing body weight on risk factors, morbidity and mortality

Risk factors	Morbidity	Mortality
Raised blood pressure	Type 2 diabetes	Increased total and cardiovascular mortality
Raised total and LDL-cholesterol	Insulin resistance
Reduced HDL-cholesterol	Coronary heart disease
Increased waist circumference	Stroke
Sleep apnoea	Osteoarthritis (knee)
Obesity hypoventilation syndrome	Cancer
Physical inactivity	Low back pain due to obesity
	Breathlessness
	Polycystic ovary syndrome Infertility
	Cholelithiasis
	Asthma (exacerbation)
	Venous thromboembolic
	Pulmonary embolism
	Inflammation
	Autonomic nervous system dysfunction

8.10 Management of obesity and overweight

Intentional weight loss in obese patients can improve or prevent many of the obesity-related risk factors for CHD. It is important for cardiovascular healthcare professionals to understand the clinical effect of weight loss and be able to implement appropriate weight-management strategies in obese patients [293]. The beneficial effects of weight reduction on the cardiovascular system are listed in Table 6 [245]. Reduction in blood pressure occurs before attainment of desirable body weight. In a recent meta-analysis blood pressure reduction were −1.05 mmHg systolic and −0.92 mmHg diastolic/kg of weight loss [294]. A Cochrane review concluded that weight reducing diets in overweight hypertensive persons can affect modest weight loss in the range of 3-9% of body weight and are probably associated with modest blood pressure decreases of roughly 3 mmHg systolic and diastolic. Weight reducing diets may decrease dosage requirements of persons taking antihypertensive medications [295].

8.11 Diet and behavioural interventions

Many different diets and behavioural interventions have been proposed for the treatment of obesity. The control of overweight is dependent upon achieving the appropriate balance between energy intake and expenditure. The dietary approaches vary in their total energy content, macronutrient composition (protein, carbohydrates and lipids), energy density and glycaemic index [293]. The low fat diet is considered the standard approach to weight reduction and has a more favourable effect on LDL-cholesterol. Total fat intake should be kept between 25 and 35% of energy. Reduction in saturated fats is the preferred target due to its effects on the lipoprotein profile. Saturated and trans FA intake should be less than 7% [162].

OPEN IN VIEWER

Table 6

Effects of weight reduction on the cardiovascular system

↓Blood volume

↓Stroke volume

↓Cardiac output

↓Pulmonary capillary wedge pressure

↓Left ventricular mass

Improvement of left ventricular diastolic dysfunction

Improvement of left ventricular systolic dysfunction

↓Resting oxygen consumption

↓Systemic arterial pressure

↓Resting heart rate

↓QTc interval

The low-carbohydrate diet has become popular and in the short-term decreases body weight greatly and also has a good effect on plasma triglycerides and HDL-cholesterol [296]. However, the long-term safety is still under investigation. Alcohol is a major source of calories and reduction may be an important part of weight control.

Behaviour modification inducing long-term lifestyle change leading to a gradual weight loss is the basis of all obesity treatment. The behaviour change process is facilitated by the use of goal-setting, self-monitoring and problem-solving techniques [297]. Studies using this approach often achieve a weight loss of 8-10% of the initial body weight at 6 months follow-up [298]. One key feature of such a programme is to discuss with the patient, in an empathic manner, both short-term and long-term goals. Furthermore, the goals should be realistic, measurable, concrete and engaging. Another key feature is self-monitoring, where the patient keeps a detailed record of his behaviours such as food intake and physical activity. Using goal-setting and self-monitoring together, individuals may easily experience satisfaction by looking at their recordings and find out that their goals have been attained. This will boost the self-efficacy for the behaviour in question. A randomized trial examining the combined effects of lifestyle modification and pharmacologic treatment (using sibutramine), showed that the combination gave a significantly larger weight loss than sibutramine alone at 1 year follow-up [299], indicating that medication may complement but should not replace efforts to achieve healthier behaviour.

According to a recent Cochrane review, behavioural and cognitive-behavioural therapy help with losing additional weight when added to diet and exercise programs [300]. Behavioural interventions have also been shown to help maintain weight loss [301].

8.12 Drug treatment of overweight

In general, the contribution of drug treatments is modest and, in the past, some products have had serious side effects.

Orlistat inhibits intestinal lipases to prevent the hydrolysis and uptake of fat. Weight loss is usually modest and gastro-intestinal disturbance may occur. It should be used with full and balanced diet.

Sibutramine enhances a feeling of satiety after food by an effect of its metabolites which inhibit noradrenaline and serotonin uptake. Both contraindications and side effects are appreciable.

Rimonabant is an endocannabinoid receptor inhibitor that appears capable of inducing a modest but sustained weight loss in combination with a calorie controlled diet. It may improve glucose tolerance, beneficially affect lipid metabolism and is associated with a modest reduction in blood pressure. Possible adverse effects on depression are being monitored. It remains to be seen if its promising effects on weight and other risk factors will translate into hard evidence of reduced cardiovascular events.

8.13 Practice points

9.1 Scientific background

9.2 Estimating physical activity

For an assessment of physical activity three different methods may be used: (i) criterion methods, for example, doubly labeled water, indirect calorimetry or direct observation, (ii) objective methods, for example, activity monitors (pedometers, accelerometers) or heart rate monitors, (iii) subjective methods such as questionnaires or activity diaries [338]. For physical fitness and exercise capacity maximal incremental exercise testing is used.

9.3 Management of physical activity

Practice point: All individuals should be professionally encouraged and supported to increase their physical activity to the level associated with the lowest risk of CVD. Although the goal is at least half an hour of physical activity on most days of the week, almost any increase in activity is associated with a variety of health benefits – a very encouraging message!

OPEN IN VIEWER

Table 7

Recommendations for physical activities

Aim

In all age groups: 30-45 min of physical activity at least five days a week

Rationale

To prevent or delay the onset of cardiovascular disease

To limit the progress of cardiovascular disease

Method

Promote daily physical exercise at school

Provide options for regular physical activity at the work site, encourage an active leisure time, e.g. brisk walking, cycling, swimming, gardening or other in/outdoor sports and hobbies

For coronary patients: participation in supervised or home-based programmes of physical training

For elderly: stimulate the maintenance of a physically active lifestyle, even in higher age groups

Result

Lower risk of cardiac mortality and morbidity

Adequate level of physical fitness, increase of VO2 max and endurance capacity

Lowering of heart rate and blood pressure

Improvement of coronary blood flow

Effect on symptoms of angina pectoris

Adaptation of the peripheral resistance

Protective effect on the sympatico-vagal balance

Reduction of overweight

Cardioprotective effect on lipid metabolism and insulin sensitivity

Effect on platelets and fibrinolysis

Healthy people should be recommended to choose enjoyable activities which fit into their daily routine, preferably for 30-45 min, 4-5 times weekly at a 60-75% of the age-adapted maximum heart rate. For patients with established CVD and for those with a high CVD 10-year mortality risk, advice must be based on a comprehensive clinical judgement, including the results of an exercise test.

In addition to improving aerobic fitness, physical activities that facilitate endurance, strength, balance and flexibility should be encouraged.

9.4 The use of exercise training in adults with CVD

Recommendations for patients with clinically established CVD have to be based on a comprehensive clinical judgement including the results of exercise testing. Patients with unstable angina pectoris should be attended to with conventional noninvasive or invasive methods before they may be included in physical training. Patients with stable angina pectoris often obtain marked subjective benefit from gradually increased and regular exercise, but their antianginal and other medical treatment should be optimal before starting such a programme. The intensity and duration should initially be set low and increased step-wise according to the limits imposed by exercise-induced symptoms. Here, the result of pretraining exercise testing will be valuable.

Patients recovering from an acute myocardial infarction or other ischaemic event and, similarly, patients following angioplasty or recovering from coronary artery bypass grafting should be recommended to start a suitable, increasing physical activity programme. Many patients will benefit from an organized rehabilitation programme provided by a multidisciplinary team. Such a programme may be available on an ambulatory basis or as an in-patient facility in specialized centres, as is the tradition in some central European countries. The rehabilitation programmes, in addition to supervised physical exercise, give a good opportunity for a comprehensive evaluation of the patient's risk factor status and for further advice and measures aimed at risk reduction.

If patients prefer to undertake their physical training programme at home, they will need clear prescriptions, encouragement and regular follow-up by their physician. Written material, audiocasettes, videos or CD-ROM disks are useful supplements to verbal advice. Devices such as heart rate monitors or pedometers maybe helpful in the follow-up of a home-based programme for physical training.

Detailed recommendations on exercise prescription and rehabilitation for cardiac patients, as well as on counselling for recreational and vocational activities have been published by the European Society of Cardiology Working Group on Rehabilitation [348], the ESC Study Group of Sports Cardiology [347], the American Association of Cardiovascular & Pulmonary Rehabilitation [349] and other experts in this field [350–352].

For patients with mild to moderate heart failure guidelines are available, issued by the ESC Working Groups on Heart Failure and on Cardiac Rehabilitation & Exercise Physiology [342]. Different forms of aerobic training have shown beneficial effects on exercise capacity, the need for rehospitalization and on quality of life, but there are as yet no data on mortality. Although there is no consensus on the optimal training regimen in CHF both dynamic interval training with moderate intensity and resistance training may be advantageous.

9.5 Promoting physical fitness in the elderly

General public health measures regarding the promotion of physical fitness apply to the elderly population as they do to younger generations. With the growing number of senior citizens these measures will have a major impact on healthcare resources. In several European countries gymnastic classes for elderly are being organized. Sports clubs, patient organizations and other commercial and noncommercial institutions provide a growing diversity of physical exercise facilities. Yet, further initiatives are recommended, especially for the large group of sedentary elderly women.

When counselling elderly persons the family doctor is recommended to assess activity regularly and advise to maintain daily physical activity on a moderate to submaximum level. The physician should be aware of the risk/benefit ratio when prescribing exercise. Low-intensity and longer duration exercise should be preferred over high intensity, short duration. Brisk walking at a pace at which a conversation still can be held (walk and talk model) is a good example of advise for the healthy elderly and carries a low risk/benefit ratio. The key-elements of activity programmes for the elderly are: a combination of endurance, strength, balance and flexibility training. Principles of behavioural change including social support, self-efficacy and positive reinforcement should be applied and programmes should start off at low intensity but gradually increase to moderate levels [353].

Older patients with signs of CVD will benefit from comprehensive rehabilitation programmes: exercise training is safe and improves strength, aerobic fitness, endurance and physical function. It will improve conventional risk factors, mental state and quality of life [354]. There are no gender differences in the outcome of training for elderly CVD patients. Resistance training may be an attractive alternative; it can be used in home-training, if transport is a limiting factor for participation in ambulatory group training.

However, as in the healthy population of all ages and among patients with established CVD the ultimate goal of all physical training programmes should be the acceptance and maintenance of a lifestyle in which efforts of regular physical exercise are rewarded by the cardiovascular as well as other benefits of general physical fitness.

Cognitive behavioural counselling enhances the effect of exercise recommendations. Physicians should therefore use techniques to increase motivation and self-efficacy. Group-based cognitive-behavioural interventions should be offered if simple recommendations fail.

9.6 Heart rate

10.1 Blood pressure as a risk factor for CVD

Elevated blood pressure has been identified as a risk factor for coronary heart disease (CHD), heart failure, stroke, peripheral arterial disease and renal failure in both men and women in a number of epidemiological studies [378–381]. Observational evidence is also available that blood pressure levels correlate negatively with cognitive function and that hypertension is associated with an increased incidence of dementia [382]. A large compilation of observational data [381] confirms that both systolic and diastolic BP show a continuous graded independent relationship with the risk of stroke and coronary events. Data involving one million individuals have indicated that death from both CHD and stroke increases progressively and linearly from BP levels as low as 115 mmHg systolic and 75 mmHg diastolic upward [383]. Increased risks are present in all age groups ranging from 40 to 89 years old. For every 20 mmHg systolic or 10 mmHg diastolic increase in BP, there is a doubling of mortality from both CHD and stroke [383].

In addition, longitudinal data obtained from the Framingham Heart Study indicated that BP values in the 130139/85-89 mmHg range are associated with more than a two-fold increase in relative risk from CVD compared with those with BP levels below 120/80 mmHg [384].

The apparently simple direct relationship between increasing systolic and diastolic BP and CV risk is confounded by the fact that systolic BP rises throughout the adult age in the vast majority of populations whereas diastolic BP peaks at about age 60 in men and 70 in women, and falls gradually thereafter [385].

This observation helps to explain why a wide pulse pressure (systolic BP-diastolic BP) has been shown in some observational studies to be a better predictor of adverse CV outcomes than either systolic or diastolic BP individually [386] and to identify patients with systolic hypertension who are at particularly high risk [387]. However, in the largest meta-analysis of observational data in one million patients in 61 studies (70% of which have been conducted in Europe) [383], both systolic and diastolic BP were independently predictive of stroke and CHD mortality and more so than pulse pressure. This meta-analysis also confirmed the increasing contribution of pulse pressure after age 55.

It has also been shown that, compared to normotensive individuals, those with an elevated blood pressure are more likely to have other risk factors for CVD such as diabetes, insulin resistance and dyslipidaemia [379, 388–390] and various types and degrees of target organ damage. Because risk factors may interact positively with each other, the overall cardiovascular risk of hypertensive patients may be high even if blood pressure is only moderately raised [104, 384].

Blood pressure can be reduced either by lifestyle interventions or by drugs. Lifestyle measures should be instituted in all patients including individuals with high-normal BP and patients who require drug treatment. The purpose is to lower BP and to control other risk factors, thereby potentially reducing the occurrence of BP-related clinical conditions. In nonhypertensive individuals, including those with high-normal BP, dietary changes that lower BP have the potential to prevent hypertension and, more broadly, to reduce BP and thereby lower the risk of BP-related clinical complications. However, lifestyle measures have not been shown to prevent CV complications in hypertensive patients, and should never delay the initiation of drug treatment unnecessarily, especially in patients at higher levels of risk, or detract from adherence to drug treatment. On the other hand, even an apparently small reduction in BP, if applied to an entire population, could have an enormous beneficial impact [102]. A substantial body of evidence strongly supports the concept that multiple nutritional factors affect BP [206]. Well-established dietary modifications that lower BP are reduced salt intake, weight loss, and moderation of alcohol consumption among those who drink. Over the past decade, increased potassium intake and dietary patterns based on the DASH diet (a diet rich in fruit, vegetables, and low-fat dairy products, with a reduced content of dietary cholesterol as well as saturated and total fat) [216] have emerged as effective strategies that also lower BP. The effects of sodium reduction on BP tend to be greater in blacks, middle-aged, and older persons and in individuals with hypertension, diabetes, or chronic kidney disease. These groups tend to have a less responsive renin-angiotensin-aldosterone system [391]. The recommended adequate sodium intake has been recently reduced from 2.4 to 1.5 g/day (65 mmol/day) [392], corresponding to 3.8 g/day sodium chloride.

A substantial and largely consistent body of evidence from observational studies and clinical trials documents that body weight is directly associated with BP. In one meta-analysis, mean systolic and diastolic BP reductions associated with an average weight loss of 5.1 kg were 4.4 and 3.6 mmHg, respectively [294]. In a subgroup analysis, BP reductions were similar for nonhypertensive and hypertensive individuals but were greater in those who lost more weight. Within-trial dose-response analyses [393, 394] and prospective observational studies [395] also document that greater weight loss leads to a greater BP reduction. Modest weight loss with or without sodium reduction can prevent hypertension in overweight individuals with high-normal BP [396], and can facilitate medication step-down and drug withdrawal [397, 398].

Several predominantly small clinical trials and meta-analyses of these trials [399–401] have documented that high-dose ω-3 polyunsaturated fatty acid (commonly called fish oil) supplements can lower BP in hypertensive individuals with BP reductions occurring at relatively high doses (≥ 3 g/day). In hypertensive individuals, average systolic and diastolic BP reductions were 4.0 and 2.5 mmHg, respectively [401].

Overall, data are insufficient to recommend an increased intake of fibre alone [402], supplemental calcium or magnesium [403, 404] as means to lower BP. Additional research is warranted before specific recommendations can be made about the amount and type of carbohydrate [405, 406] to affect BP.

Epidemiological studies suggest an inverse relationship between habitual physical activity and BP. Higher levels of physical activity and greater fitness are associated with a reduced incidence of hypertension [407]. A meta-analysis of controlled interventional trials concluded that adequate dynamic physical training contributes usefully to blood pressure control. In normotensive individuals, the training-induced decrease in BP averaged 2.6/1.8 mmHg, and in hypertensive individuals, 7.4/5.8 mmHg [408, 409]. The BP-lowering effects of exercise are most pronounced in hypertensive individuals engaging in endurance exercise with BP decreasing approximately 5-7 mmHg after an isolated exercise session (acute) or following exercise training (chronic). Moreover, BP is reduced for up to 22 h after an endurance exercise bout, with the greatest decreases among those with the highest baseline BP. In a recent meta-analysis of randomized controlled trials, resistance training at moderate intensity was found to decrease BP by 3.5/3.2 mmHg [410]. For hypertensive individuals, an exercise program that is primarily aerobic-based with adjunctive resistance training is recommended [407]. The evidence is limited regarding frequency, intensity, and time recommendations; the antihypertensive effect appears to occur at relatively low duration and intensity.

Systolic BP during bicycle exercise (no systematic studies using a treadmill are available) has been proposed as a more sensitive indicator of CV risk and risk of developing hypertension [411]. A rise in exercise systolic BP to >200 mmHg during the first 6 min of bicycle exercise predicts a doubling of CVD incidence in middle-aged men [412]. If the exercise-induced rise in cardiac output is impaired in hypertensive individuals, exercise BP no longer carries an independent prognostic power [413].

Smoking causes an acute increase in BP, heart rate, and circulating catecholamines persisting for more than 15 min after smoking one cigarette. Paradoxically, several epidemiological studies have found that BP levels among cigarette smokers were the same as or lower than those of nonsmokers [414, 415]. A few studies using ABPM (ambulatory blood pressure monitoring) have shown that both untreated hypertensive and normotensive smokers present higher daytime BP than nonsmokers [416–419]. Although any independent chronic effect of smoking on BP is small [420] and smoking cessation does not lower BP [421], total cardiovascular risk is greatly increased by smoking [422]. Patients should be strongly counselled to quit smoking to reduce overall CV risk.

Combinations of two or more lifestyle modifications can achieve better results [215].

It should be emphasized that the size of all lifestyle intervention trials has been too small and their duration too short to provide evidence of the effect of lifestyle changes on cardiovascular morbidity and mortality. Furthermore, adherence to lifestyle changes out of the context of controlled trials has been shown to be poor, particularly with regard to weight loss whose maintenance represents a universal problem in the long-term [423]. However, the TOHP study [396] has shown that even when body weight is regained, blood pressure may remain lower than in the control group. This finding suggests that some degree of weight loss, even if not sustained beyond 6 months, confers benefit [394]. Also, in responsive and compliant individuals, lifestyle changes may (i) decrease the number and doses of antihypertensive drugs necessary to control blood pressure, (ii) make it unnecessary to restart medication after effective drug treatment has been stopped, and (iii) reduce the overall cardiovascular risk profile. This makes this approach mandatory under all circumstances.

Large-scale randomized controlled trials performed mostly in Caucasian populations have conclusively demonstrated that, in hypertensive individuals a blood pressure reduction by antihypertensive drugs substantially reduces cardiovascular morbidity and mortality [424–428]. It should be noted that most of the hypertension trials involved patients over 55 years of age with overt vascular disease and that there is an alarming absence of data in younger (>55 years old) individuals [429].

The trials have also provided evidence that the benefit:

Evidence from placebo-controlled and comparative trials also makes it clear that cardiovascular protection can be obtained by treatments based on a variety of antihypertensive drug classes, that is, diuretics, β-blockers, ACE inhibitors, calcium antagonists and angiotensin II antagonists. This presumably means that the protection is due, to a substantial degree, to blood pressure lowering per se [4].

The benefit of β-blockers compared with that of other antihypertensive agents has recently been questioned on the basis of the results of two large randomized trials, the LIFE study [442] and the ASCOT study [443], both of which showed superiority of an angiotensin receptor antagonist and a calcium antagonist, respectively, over therapy initiated by a β-blocker as far as stroke (LIFE) or stroke and mortality (ASCOT) were concerned. These two large trials have strongly influenced a recent meta-analysis [444], which concluded that β-blocker-initiated therapy is inferior to others in stroke prevention, but not in prevention of myocardial infarction and reduction in mortality. On the basis of a similar meta-analysis, NICE in the United Kingdom has advised to use β-blockers only as fourth-line antihypertensive agents [429]. These conclusions must be considered with critical caution. Both the LIFE and the ASCOT studies were characterized by a design that led to early use of combination therapy, so that the vast majority of the patients randomized to a β-blocker-based therapy actually received a β-blocker-thiazide combination. A recent meta-analysis shows that, when compared to placebo, β-blocker-based therapy did indeed reduce stroke significantly [445]. Part of the β-blocker-thiazide combination reported in ASCOT may be due to a smaller reduction of blood pressure [443], particularly of central blood pressure [446] that occurred in this trial with this therapeutic regimen.

β-Blocker-thiazide combinations have nevertheless been consistently associated with metabolic disturbance and new-onset diabetes and may have specific contraindications in patients prone to diabetes. In any case, the above meta-analyses of β-blocker-initiated trials [444, 445] well illustrate the difficulties inherent in many recent trials in which combination therapy hinders the attribution of either benefits or harms to individual compounds.

Following a myocardial infarction, BP elevation is associated with an increased risk of reinfarction and death [447–449]. No randomized controlled trial evidence is available on the effect of BP lowering per se under these circumstances. However, use of β-blockers, ACE inhibitors and nondihydropyridine calcium antagonists immediately after an acute myocardial infarction has resulted in a secondary cardioprotective effect together with a modest BP reduction [450–452]. This has been seen both in individuals with elevated and normal BP. Furthermore, antihypertensive treatment by several drug classes has been found to prevent cardiovascular disease in normotensive and hypertensive patients with a more distant history of myocardial infarction [453]. Finally, recent trials have demonstrated that clinically stable normotensive and hypertensive patients with a history of cerebrovascular disease show a marked decrease in their otherwise elevated incidence of stroke recurrence if blood pressure is reduced by ACE inhibitors in combination with diuretics [454], with a concomitant decrease in the incidence of myocardial infarction. This suggests use of antihypertensive drugs for secondary coronary and cerebrovascular prevention, even when blood pressure is not elevated. Treatment should be given according to guidelines for primary prevention but BP should be lowered slowly and carefully because tissue necrosis, atherosclerotic plaques and cardiac hypertrophy may make coronary and cerebral blood pressure flow auto-regulation less effective in preserving organ perfusion when perfusion pressure is reduced [455, 456]. In addition, the HOPE study [457] suggested that treatment of high risk people with an ACE inhibitor may have benefits beyond blood pressure control.

Long-term observational data [458] provide evidence that, in hypertensive patients in whom treatment effectively controls BP, coronary, cerebrovascular and overall CV morbidity remains higher than that of normotensive controls. This may be accounted for by factors such as irreversible organ damage at the time treatment is started pointing to the need for earlier identification and correction of blood pressure elevation. Research efforts, however, currently also focus on the possibility that greater cardiovascular protection may be achieved by (i) antihypertensive drugs with direct organ protective properties that might complement the protection due to the blood pressure reduction [453], (ii) multiple drug treatments that more comprehensively address the overall cardiovascular risk profile, and (iii) more aggressive blood pressure reductions below 140/90 mmHg. To date, trial evidence that for a given BP reduction some antihypertensive agents are more protective than others is limited to the greater nephroprotective effects of drugs primarily acting on the renin-angiotensin system (ACE inhibitors and angiotensin II antagonists) [459–461] because, with few exceptions [442, 462], comparisons between drugs have not shown differences in primary cardiovascular end points and no substantial advantage of one class over another had emerged from their meta-analysis [463, 464]. Similar results were obtained in a recent meta-analysis performed by the same collaborative group in patients with and without diabetes showing a comparable reduction of total major CV events [465]. Another meta-analysis comparing the effects of ACE inhibitors and angiotensin II antagonists with other antihypertensive drugs yielded a relative risk of 0.71 for doubling of creatinine (95% CI of 0.49-1.04) and a small benefit on end-stage renal disease (relative risk, 0.87; 0.75-0.99). In patients with diabetes, additional renoprotective action of the substances beyond lowering blood pressure remains unproven [466].

Several large-scale studies have shown statins to be very effective in primary and secondary prevention and also in patients with hypertension and antihypertensive treatment [63, 467]. Two trials – ALLHAT [468] and ASCOT [469] – have evaluated the benefits associated with the use of statins, specifically in patients with hypertension. ALLHATcompared the effect of 40 mg/day pravastatin with usual care in over 10 000 patients, 14% of whom had established vascular disease. The differential effect of pravastatin on total and LDL-cholesterol (11 and 17%, respectively) was smaller than expected due to extensive statin use in the usual care group and was associated with a modest, nonsignificant 9% reduction in fatal CHD and nonfatal myocardial infarction, and a 9% reduction in fatal and nonfatal stroke. There was no impact on all-cause mortality, which was the primary end point of the trial. By contrast, the results of ASCOT, which also included over 10 000 hypertensive patients, showed a 36% reduction in the primary end point of total CHD and nonfatal myocardial infarction, and a 27% reduction in fatal and nonfatal stroke associated with the use of atorvastatin 10 mg/day compared with placebo in patients with total cholesterol less than 6.5 mmol/l [469]. Based on ASCOT data, well controlled high risk hypertensive individuals may receive a modest additional benefit from the use of a statin.

Evidence has been obtained, on the other hand, that in hypertensive patients in whom treatment provides an adequate blood pressure control, the addition of acetylsalicylic acid (ASA) at a low daily dose (75 mg) further decreases (−35%) the incidence of myocardial infarction [470], the benefit being particularly evident in males [471] and in the subgroup with renal damage [472]. However, an increase in rates of nonfatal major bleeds was also shown and therefore, ASA should not be recommended in hypertension with low total CV risk and poor BP control.

There is also evidence that in patients with type 2 diabetes mellitus or with diabetic nephropathy, aiming at blood pressure values well below the traditional ones is associated with a lower incidence of cardiovascular and renal events, respectively [473–475]. The latter is the case particularly in the presence of marked proteinuria. Furthermore, in diabetic and nephropathic [476, 477] patients on-treatment blood pressure values less than 130/80 mmHg or lower (>125/75 mmHg in patients with proteinuria >1 g/24 h) should be achieved using all available drugs with evidence of antihypertensive efficacy and safety [478, 479].

10.2 Blood pressure measurements

The large physiological variations in blood pressure [480] mean that, to diagnose hypertension, blood pressure should be measured in each individual several times on several separate occasions. If systolic and/or diastolic blood pressure is only slightly elevated, repeated measurements should be made over a period of several months to achieve an acceptable definition of the individual's ‘usual’ blood pressure and to decide about initiating drug treatment. If systolic and/or diastolic blood pressure is more markedly elevated, repeated blood pressure measurements are required within a shorter period of time (weeks or days) in order to make treatment decisions. This is also the case if the blood pressure elevation is accompanied by evidence of endorgan damage, associated clinical conditions, and/or by the concomitance of other cardiovascular risk factors that markedly increase overall cardiovascular risk. Repeated blood pressure measurements on several occasions are necessary to identify the relatively large number of persons in whom blood pressure elevation disappears following the first few visits. These individuals may need blood pressure measurements more frequently than the general population but drug treatment does not appear to be necessary because their cardiovascular risk is probably low [481].

10.3 Isolated office or white-coat hypertension

In some patients, office BP is persistently elevated while daytime or 24-h BP falls to within the normal range. This condition is widely known as ‘white-coat hypertension’ [506], although the more descriptive and less mechanistic term ‘isolated office (or clinic) hypertension’ is preferable because the office-ambulatory BP difference does not correlate with the office BP elevation induced by the alerting response to the doctor or the nurse, that is, the true ‘white-coat effect’ [507]. Regardless of the terminology, evidence is now available that isolated office hypertension is not infrequent (about 15% in the general population) [508] and that it may account for a noticeable fraction of individuals in whom hypertension is diagnosed [494, 497, 509]. There is also evidence that, in individuals with isolated office hypertension, CV risk is less than in individuals with both office and ambulatory BP elevations [508]. Sustained hypertension may develop in some patients with white-coat hypertension, and the risk of stroke may increase [510]. Several, although not all studies, however, have reported this condition to be associated with a prevalence of organ damage and metabolic abnormalities greater than that of normal individuals, which suggests that it may not be an entirely innocent phenomenon [497]. This was also confirmed more recently by the PAMELA study showing an increase in both CV and all-cause mortality in white-coat hypertensive individuals [511].

Physicians should diagnose isolated office hypertension whenever office BP is ≥ 140/90 mmHg at several visits while 24-h and day ambulatory BP are less than 125-130/80 and 130-135/85 mmHg, respectively. Diagnosis can also be based on home BP values (average of several days’ readings >135/85 mmHg). Identification should be followed by the search for metabolic risk factors and target organ damage (TOD). Drug treatment should be instituted when there is evidence of organ damage or a high CV risk profile. Lifestyle changes and a close follow-up should, however, be implemented in all patients with isolated office hypertension in whom the doctor elects not to start pharmacological treatment.

10.4 Masked hypertension or isolated ambulatory hypertension or ‘reversed white-coat’ condition

Normal BP in the office and elevated blood pressures elsewhere (e.g. at work or at home) is called masked hypertension or isolated ambulatory hypertension [495, 497, 505, 512]. These individuals have been shown to display a greater than normal prevalence of TOD [497, 513] and may have a greater CV risk than truly normotensive individuals [495, 505, 511]. Alcohol, tobacco, and caffeine consumption and physical inactivity outside the office/clinic may contribute to this phenomenon. The prevalence of masked hypertension in treated hypertensive individuals is about 10% [514] and is somewhat greater in the general population [515]; in the PAMELA study 1 in seven to eight normotensive patients had it [497]. The clinic BP of patients with masked hypertension may underestimate the risk of CV events. A study of patients with treated hypertension showed about one third of those seen in a hypertension clinic had masked hypertension over a 5-year follow-up period, and their relative risk of CV events was 2.28 as compared with patients whose BP was adequately controlled according to the criteria for both clinic BP and ambulatory BP [516]. Other studies have shown masked hypertension in patients with untreated hypertension and often in those with undiagnosed hypertension, it is associated with an increased rate of TOD [513] and adverse prognosis [517]. Masked hypertension may be suspected on the basis of high BP readings taken at home, and one study has shown that masked hypertension diagnosed solely on the basis of home recordings is associated with increased mortality [505].

In patients with an acute myocardial infarction who have been treated for hypertension before their infarction, blood pressure may remain at much lower levels, or even return to normotensive values, for months or years without continuing antihypertensive treatment [518]. This observation preceded the widespread use of β-blocking drugs after myocardial infarction. In such instances, the blood pressure level has to be measured properly to detect whether and when hypertensive values are regained and effective antihypertensive treatment should be restarted without delay.

10.5 Control of arterial hypertension

Guidelines on the management of hypertension vary slightly in their definitions of hypertension and its subdivision into further blood pressure categories [519, 520]. As stated in the 1996 World Health Organization Expert Committee report on hypertension control [521], all definitions of hypertension are by necessity arbitrary because the risk of cardiovascular disease increases continuously with rising blood pressure. In epidemiologic studies, there is a continuous relationship with cardiovascular risk down to systolic and diastolic levels of 115-110 and 75-70 mmHg, respectively [381, 383]. The dividing line between ‘normotensive’ and ‘hypertensive’ individuals can only be determined operationally by intervention trials demonstrating at which blood pressure levels treatment is beneficial.

OPEN IN VIEWER

Table 8

Definition and classification of blood pressure levels

Category	Systolic		Diastolic
Optimal	>120	And	< 80
Normal	120-129	and/or	80-84
High normal	130-139	and/or	85-89
Grade 1 hypertension	140-159	and/or	90-99
Grade 2 hypertension	160-179	and/or	100-109
Grade 3 hypertension	≥ 180	and/or	≥ 110
Isolated systolic	≥ 140	And	>90

Isolated systolic hypertension should be graded (1, 2, 3) according to systolic blood pressure values in the ranges indicated, provided that diastolic values are >90 mmHg. Grades 1, 2 and 3 correspond to classification of mild, moderate and severe hypertension, respectively. These terms have now been omitted to avoid confusion with quantification of total cardiovascular risk.

The classification of hypertension used in the 2003 and 2007 ESH-ESC guidelines has been retained [4, 520] (Table 8). Isolated systolic hypertension should be graded to the same systolic blood pressure values indicated for systolic-diastolic hypertension. However, the association with a low diastolic blood pressure (e.g. 60-70 mmHg) should be regarded as an additional risk.

The decision to start pharmacological treatment, however, depends not only on the blood pressure level but also on total cardiovascular risk, which calls for a proper history, physical examination and laboratory examination to identify (i) the presence of clinically established cardiovascular or renal disease, (ii) the coexistence of other cardiovascular risk factors, and (iii) the presence of subclinical cardiovascular disease or end-organ damage. The presence of clinically established cardiovascular or renal disease (myocardial infarction, angina pectoris, heart failure, coronary revascularization, transient is-chaemic attacks, stroke, renal insufficiency or overt proteinuria, peripheral arterial disease, advanced retinopathy, etc.) dramatically increases the risk of subsequent cardiovascular events regardless of the blood pressure level. This is also the case for the association of hypertension and other cardiovascular risk factors such as diabetes (Table 9).

Owing to the importance of target organ damage as an intermediate stage in the continuum of vascular disease and as a determinant of overall cardiovascular risk, signs of organ involvement should be looked for carefully.

OPEN IN VIEWER

Table 9

Factors influencing prognosis in hypertension

Risk factors	Target organ damage	Diabetes mellitus	Established CV or renal disease
•Systolic and diastolic BP levels	•Electrocardiographic LVH	•Fasting plasma glucose ≥ 7.0 mmol/l (126 mg/dl)	•Cerebrovascular disease: ischaemic stroke; cerebral haemorrhage; transient ischaemic attack
•Levels of pulse pressure (in the elderly)	(Sokolow-Lyons >38 mm; Cornell >2440 mm × ms)	or
•Age (M >55 years; W >65 years)	Or	•Postload plasma glucose >11.0 mmol/l (198 mg/dl)
•Smoking	•Echocardiographic LVHa (LVMI M ≥ 125 g/m2, W ≥ 110 g/m2)		•Heart disease: myocardial infarction; angina; coronary revascularization; heart failure
•Dyslipidaemia
TC >5.0 mmol/l (190 mg/dl)
or	•Carotid wall thickening (IMI ≥ 0.9 mm) or plaque
LDL-C >3.0 mmol/l (115 mg/dl)	•Carotid-femoral pulse wave velocity >12 m/sec
or	•Ankle/Brachial BP index >0.9
HDL-C: M >1.0 mmol/l (40 mg/dl), W < 1.2 mmol/l (46 mg/dl)
or			•Renal disease: diabetic nephropathy; renal impairment (serum creatinine M>133, W>124 mmol/l) proteinuria (>300 mg/24 h)
TG >1.7 mmol/l (150 mg/dl)	•Slight increase in plasma creatinine:
•Fasting plasma glucose 5.6-6.9 mmol/l (100-125 mg/dl)	M: 115-133 mmol/l (1.3-1.5 mg/d); W: 107-124 mmol/l (1.2-1.4 mg/dl)
•Abnormal glucose tolerance test	•Low estimated glomerular filtration rateb
•Abdominal obesity (waist circumference >102 cm (M), 88 cm (W))	(>60 ml/min/1.73 m2) or creatinine clearancec (>60 ml/min)		•Peripheral artery disease
•Family history of premature CV disease (M at age >55 years; W at age < 65 years)	•Microalbuminuria 30-300 mg/24h or albumin-creatinine ratio: ≥ 22(M); or ≥ 31 (W) mg/g creatinine		•Advanced retinopathy: haemorrhages or exudates, papilloedema

Legend M, men; W, women; CV, cardiovascular disease; IMT, intima-media thickness; BP, blood pressure; TG, triglycerides; C, cholesterol. ^aRisk maximal for concentric LVH (left ventricular hertrophy): increased LVMI (left ventricular mass index) with a wall thickness radius ratio ≥ 0.42. ^bMDRD formula. ^cCockcroft-Gault formula.

Electrocardiography should be part of routine assessment of hypertensive patients. Its sensitivity in detecting left ventricular hypertrophy is low, but hypertrophy detected by the Sokolow-Lyons index (SV₁ + RV_5-6 >38 mm) or by the Cornell voltage QRS duration product (> 2440 mm × ms) is an independent predictor of cardiovascular events [522]. Use of ECG hypertrophy as a marker of cardiac damage as well as a tool documenting LVH regression, which may also be associated with a reduced incidence of new-onset atrial fibrillation appears to be valuable, at least in patients aged more than 55 years [523].

Echocardiography is more sensitive than electrocardiography in diagnosing left ventricular hypertrophy [524] and predicting cardiovascular risk [525], and may help in more precise stratification of the overall risk and in directing therapy [390].

Carotid ultrasound with measurement of intima-media thickness (IMT) or the presence of plaques has been shown to predict both stroke and myocardial infarction [526–530]. The relationship between carotid IMT and cardiovascular events is a continuous one but for the common carotid arteries an IMT >0.9 mm can be taken as a conservative estimate of existing alterations. Ultrasound scans limited to the common carotid arteries (an infrequent site of atherosclerosis) are likely to detect vascular hypertrophy only, whereas assessment of atherosclerosis also requires scanning of bifurcations and/or internal carotids where plaques are more frequent. Presence of a plaque can be identified by an IMT >1.3 or 1.5 mm or by a focal increase in thickness by 0.5 mm or by 50% of the surrounding IMT value [531, 532]. There is evidence that, in untreated hypertensive individuals without target organ damage at routine examinations, these alterations are common and, thus, carotid ultrasound may often detect vascular damage and make risk stratification more precise [390].

Evidence of arterial damage may also be suggested by an ankle-brachial blood pressure index less than 0.9, using a continuous wave Doppler unit and a blood pressure manometer. A low ankle-brachial index indicates advanced atherosclerosis [533, 534], whereas carotid IMT measurements are able to detect earlier changes [530]. Nevertheless, a reduced ankle-brachial index predicts further development of angina, myocardial infarction, congestive heart failure, need for coronary bypass surgery, stroke, carotid and peripheral artery surgery [535–539] and, in patients with multivessel coronary disease, it confers an additional risk [540].

A large body of evidence has been collected over the past 10 years on large artery stiffening and the wave reflection phenomenon, which have been identified as the most important pathophysiological determinants of isolated systolic hypertension and pulse pressure increases [541]. Measurement of carotid-femoral pulse wave velocity provides a comprehensive noninvasive assessment of arterial stiffness, which is simple and accurate enough to be considered for diagnostic purposes [542]. This measure has been shown to have an independent predictive value for all-cause and cardiovascular morbidity, coronary events and strokes in patients with uncomplicated essential hypertension [543–545]. Although the relationship between aortic stiffness and events is continuous, a threshold more than 12 m/s has been suggested as a conservative estimate of significant alterations of aortic function in middle-aged hypertensive individuals. Though a wider clinical use of pulse wave velocity measurement may add further precision to assessment of arterial damage, availability of the technique is largely limited to research centres.

The diagnosis of hypertension-induced renal damage is based on the finding of a reduced renal function and/or the detection of elevated urinary excretion of albumin. Renal insufficiency is now classified according to the estimated glomerular filtration rate calculated using the MDRD formula that requires age, gender, race and serum creatinine [546]. Values of estimated glomerular filtration rate below 60 ml/min/1.73 m² indicate chronic renal disease stage 3, while values below 30 and 15 ml/min/1.73 m² indicate chronic renal disease stages 4 and 5, respectively [547]. Another option is the Cockcroft-Gault formula estimating creatinine clearance and based on age, gender, body weight, and serum creatinine. This formula is valid in the range greater than 60 ml/min, but it overestimates creatinine clearance in chronic kidney disease stages 3-5 [548]. Both formulas help to detect mildly impaired renal function, particularly if serum creatinine values are still within the normal range [549].

While an elevated serum creatinine concentration or a low estimated glomerular filtration rate (or creatinine clearance) points to a reduced rate of glomerular filtration, an increased rate of urinary albumin or protein excretion points to a derangement in the glomerular filtration barrier. Microalbuminuria (urinary albumin excretion from 30 to 300 mg/24 h) has been shown to predict the development of overt diabetic nephropathy in both type 1 and type 2 diabetic individuals [550], while the presence of overt proteinuria (> 300 mg/24 h) generally indicates established renal parenchymatous damage [551]. In both diabetic and nondiabetic hypertensive patients, microalbuminuria, even below the currently used threshold values [552], has been shown to predict cardiovascular events [553–561], and a continuous relationship between cardiovascular, as well as noncardiovascular, mortality and urinary protein/creatinine ratios ≥ 3.9 mg/g in men and 7.5 mg/g in women has been reported in several studies [560]. Microalbuminuria can be measured from spot urine samples (24-h or nighttime urine samples are discouraged due to the inaccuracy of urinary sample collection) by indexing the urinary albumin concentration to the urinary creatinine concentration [547]. Classic dipstick tests detect albuminuria above 300 mg/g creatinine and the ‘microalbuminuric’ dipstick test above 30 mg/g creatinine. A sensitive dipstick for the lower range of low grade albuminuria is not yet available.

In conclusion, the finding of an impaired renal function in a hypertensive patient is frequent and constitutes a very potent predictor of future cardiovascular events and death even in treated patients [472, 554, 562–565]. Therefore, it is recommended that glomerular filtration rate is estimated, and the presence of urinary protein (by dipstick) sought in all hypertensive patients. In dipstick-negative patients, low grade albuminuria should also be searched for in spot urine using one of the validated commercial methods.

The coexistence of other cardiovascular risk factors (smoking, increased plasma cholesterol, family history of premature cardiovascular disease) also greatly adds to the risk associated with a mild blood pressure elevation (see SCORE risk chart) [105].

ESC guidelines (2007) [4] suggest the following tests to be performed routinely in hypertensive patients: fasting plasma glucose, serum total cholesterol, serum HDL-cholesterol, fasting serum triglycerides, serum potassium, serum uric acid, serum creatinine, estimated creatinine clearance (Cockcroft-Gault formula) or estimated glomerular filtration rate (MDRD formula), haemoglobin and haematocrit, urinalysis (complemented by micro-albuminuria dipstick test and sediment examination), and electrocardiogram, whereas echocardiography, carotid ultrasound, ankle-brachial index, fundoscopy and measurement of pulse wave velocity are listed as tests to be considered. If fasting plasma glucose is more than 5.6 mmol/l (∼ 100 mg/dl), a glucose tolerance test is recommended. Blood pressure measurement at home or 24 h ambulatory blood pressure monitoring are also part of recommended tests.

10.5.1 Who to treat?

The decision to start antihypertensive treatment depends on systolic and diastolic blood pressure, as classified in Table 8, and on total cardiovascular risk as estimated from the SCORE charts (Figs 3–6). However, in hypertensive patients, prognosis is also affected by the presence or absence of target organ damage, diabetes mellitus, and established CV or renal disease (Table 9). All patients in whom repeated blood pressure measurements show grade 2 or 3 hypertension (i.e. systolic values ≥ 160 or diastolic values ≥ 90 mmHg) are candidates for antihypertensive treatment because a large number of placebo-controlled trials have conclusively demonstrated that, in patients with these blood pressure values, blood pressure reduction lowers cardiovascular morbidity and mortality [434, 464, 566, 567]. The benefit may be modest in those at low CVD risk. Benefits of drug treatment should be weighed against side effects, cost, use of medical resources and turning healthy people into ‘patients'.

Evidence for the benefit from treating grade 1 hypertensive individuals is admittedly scanter, as specific trials have not addressed this issue. However, the recent finding of the FEVER study on the protective effect of lowering systolic blood pressure to less than 140 rather than more than 140 mmHg even in hypertensive patients at moderate risk [568] lends support to the recommendation to consider antihypertensive interventions when systolic blood pressure is ≥ 140 mmHg.

In all grade 1 to 3 hypertensive individuals, lifestyle counselling should be provided after hypertension is diagnosed, while promptness in the initiation of pharmacological therapy depends on the level of total cardiovascular risk. In the high risk hypertensive individuals enrolled into the VALUE study, the treatment arm with blood pressure control achieved later showed a trend towards more cardiovascular events [569]. Therefore, in hypertensive patients with established CV or renal disease, TOD or diabetes, the acceptable time delay to assess the results of lifestyle changes is shorter than indicated in the previous guidelines [520]. Drug treatment should be initiated promptly in grade 3 hypertension, as well as in grade 1 and 2 hypertensive individuals with high or very high total cardiovascular risk (i.e. in hypertensive patients with established CV or renal disease, target organ damage or diabetes). In grade 1 or 2 hypertensive individuals with moderate total cardiovascular risk, drug treatment may be delayed for several weeks and, in grade 1 hypertensive individuals without any other risk factor, for several months. However, even in these patients, lack of blood pressure control after a suitable period of lifestyle measures should lead to instituting drug treatment in addition to lifestyle measures. Recommendations for blood pressure management are summarized in Table 10.

When initial blood pressure is within the high normal range (130-139/85-89 mmHg), the decision on drug intervention depends heavily on total cardiovascular risk.

OPEN IN VIEWER

Table 10

Management of total CVD risk – blood pressure

Management of total CVD risk-BLOOD PRESSURE
In all cases, look for and manage all risk factors. Those with established CVD, diabetes or renal disease are at markedly increased risk and a BP of >130/80 is desirable if feasible. For all other people, check SCORE risk. Those with target organ damage are managed as ‘increased risk'
SCORE CVD risk	Normal >130/85	High Normal 130-139/ 85-89	Grade 1 140-159/ 90-99	Grade 2 160-179/ 100-109	Grade 3 ≥ 180/110
Low >1%	Lifestyle advice	Lifestyle advice	Lifestyle advice	Drug Rx if persists	Drug Rx
Mod 1-4%	Lifestyle advice	Lifestyle advice	+ consider Drug Rx	Drug Rx if persists	Drug Rx
Increased 5- 9%	Lifestyle advice	+ consider Drug Rx	Drug Rx	Drug Rx	Drug Rx
Markedly increased ≥ 10%	Lifestyle advice	+ consider Drug Rx	Drug Rx	Drug Rx	Drug Rx

In case of diabetes or a history of cerebrovascular or coronary disease, recent randomized trials [12, 570–573] have shown that antihypertensive treatment is associated with a reduction in cardiovascular fatal and nonfatal events although no benefit of blood pressure lowering was reported in two other trials in coronary patients. A reduction of cardiovascular events was only seen when initial blood pressure was within the hypertensive range [574, 575]. Evidence is also available that, in diabetic patients with increased protein excretion, reductions in blood pressure to very low values (>125/75 mmHg) are associated with reductions in urinary albumin excretion or proteinuria as well as with a reduced rate of progression to more severe proteinuric states. This is also the case when initial blood pressure values show a borderline or zero elevation, and drugs with a direct antiproteinuric effect such as blockers of the renin-angiotensin system are used [466, 573, 576–578]. This justifies the recommendation to start antihypertensive drug administration (together with intense lifestyle changes) even in patients with blood pressure in the high normal (and sometimes normal) range, with associated cardiovascular disease or diabetes.

Whether a similar therapeutic approach (i.e. intense lifestyle changes combined with antihypertensive drug treatment) may also benefit individuals with high normal blood pressure who are at high total risk is uncertain. It should be emphasized that prospective observational studies have demonstrated that individuals with high normal blood pressure have a higher incidence of cardiovascular disease compared to individuals with normal or optimal blood pressure [381, 383, 384]. Furthermore, the risk of developing hypertension is higher in individuals with high normal than in those with normal or optimal blood pressure, with an additional increase in risk when, as often occurs, concurrent multiple risk factors and the metabolic syndrome are present [579–581]. Lifestyle measures and close blood pressure monitoring should be the recommendation for individuals with normal blood pressure who are at low or moderate risk.

10.6 Blood pressure target in the general hypertensive population

The 2003 ESH-ESC Guidelines [520], while recommending to lower blood pressure below 140/90 mmHg in all hypertensive individuals, admitted that this was only a prudent recommendation, since trial evidence of the benefit of achieving this goal was limited to patients with diabetes or preexisting cardiovascular disease, and to a post hoc analysis of the HOT trial data [470], indicating the lowest event incidence to be at blood pressures of about 138/83 mmHg. In addition to the evidence reviewed in the 2003 guidelines [520], further indirect evidence supporting a blood pressure goal less than 140 mmHg has been provided by post hoc analyses of the VALUE and INVEST trials. In the VALUE study [663], hypertensive patients whose blood pressure was ‘controlled’ by treatment (>140/90 mmHg) had a significantly lower incidence of stroke, myocardial infarction, and heart failure as well as cardiovascular morbidity and mortality than those remaining ‘uncontrolled', independent of the antihypertensive regimens to which they were allocated. Lower rates of nonfatal and fatal cardiovascular events have also been reported in ‘controlled’ versus ‘uncontrolled’ hypertensive patients of the INVEST study [664]. All this is consistent with what has been reported in studies in hypertensive patients followed in the setting of clinical practice, with those achieving blood pressure values less than 140/90 mmHg showing cardiovascular morbidity and mortality rates much lower than those treated but uncontrolled [665].

On the basis of current evidence it can be recommended that in those who qualify for drug treatment, blood pressure be lowered at least to below 140/90 mmHg in all hypertensive patients and that lower values be pursued, if tolerated, in higher risk persons.

10.7 Blood pressure targets in diabetic patients

In order to maximize cardiovascular protection in diabetic patients, it has been recommended that antihypertensive treatment should be more intense and a goal blood pressure of less than 130/80 mmHg has been proposed. There is very solid evidence of a beneficial effect (reduction in macrovascular and microvascular complications) of a greater versus a smaller blood pressure reduction in type 2 diabetic patients as demonstrated by the HOT and UKPDS trials [470, 666] and confirmed by the ABCD studies [576, 667]. A recent meta-analysis of available trials in diabetic patients has calculated a reduced incidence of cardiovascular events (particularly stroke) with more versus less intense treatment, for a between-group difference in systolic and diastolic blood pressure averaging 6.0 and 4.6 mmHg, respectively [465]. Nevertheless, evidence of the benefit of the strict goal of less than 130/80 mmHg is more limited. Several randomized trials have shown the benefit of reducing diastolic blood pressure to values very close to or even below 80 mmHg [470, 573, 576, 667], but very few data are available on the beneficial effect of systolic blood pressure targets less than 130 mmHg. However, (i) in the ABCD studies [576, 667] in diabetic hypertensive or normotensive patients, achieved systolic blood pressure values of 132 and 128 mmHg, respectively, were associated with lower incidence of outcomes (total mortality and stroke, respectively) than in the groups with slightly less rigorous blood pressure control (systolic blood pressure of 138 and 137 mmHg, respectively) and (ii) a prospective observational study within the UKPDS programme has found a significant relationship between follow-up systolic blood pressure and incidence of macrovascular and microvascular complications in diabetic patients, with a continuous increment in complications for values more than 120 mmHg [668]. Episodes of postural hypertension should be avoided, particularly in diabetic patients.

10.8 Blood pressure targets in high or very high risk patients

Data favouring lower blood pressure targets in patients whose high risk condition is due to factors other than diabetes are of variable strength. The most clear evidence concerns patients with previous stroke or transient ischaemic attack. In the PROGRESS study [570], patients with a history of cerebrovascular disease in whom treatment reduced blood pressure from 147/86 to 138/82 mmHg showed a 28% reduction in stroke recurrence and a 26% reduction in the incidence of major cardiovascular events compared with placebo, in which the blood pressure reduction was negligible. There were substantial cardiovascular benefits also in normotensive patients in whom on-treatment values were reduced to 127/75 mmHg. Furthermore, in a recent post hoc analysis of the PROGRESS data, a progressive reduction in the incidence of stroke recurrence (particularly haemorrhagic stroke) has been reported until systolic blood pressure values of about 120 mmHg [669]. Lower levels of evidence are available for other high risk groups. In a post hoc subgroup analysis of the HOT study [670], greater reductions in diastolic and systolic blood pressure (82 versus 85 mmHg and 142-145 versus 145148 mmHg) were associated with a greater benefit in patients with a high or very high total cardiovascular risk (50% of the HOT population), but not in patients at a lower level of risk. In placebo-controlled trials in myocardial infarction survivors, administration of β–blockers or ACE inhibitors [629, 671] reduced the incidence of recurrent myocardial infarction and death even when blood pressure was normal.

Most placebo-controlled trials in patients with angina or coronary heart disease [571, 572, 574] have provided evidence of a reduced incidence of cardiovascular events by bringing blood pressure targets to rather low levels (EUROPA: 128/78 rather than 133/80 mmHg; ACTION hypertensives: 137/77 rather than 144/81 mmHg; CAMELOT: 124/76 rather than 130/77 mmHg) although, in another trial in anginal patients, similar blood pressure targets (129/74 mmHg rather than 132/76 mmHg) have provided no further benefit [575].

Therefore, similar targets should be adopted in individuals with a history of cerebrovascular disease and can at least be considered in patients with coronary disease.

There are no sufficient cardiovascular outcome data upon which to recommend a lower target blood pressure in patients with nondiabetic renal disease, but sufficient, though not conclusive evidence suggests that values lower than 130/80 mmHg may help in preserving renal function, especially in the presence of proteinuria.

11.1 Lipids and lipoproteins as risk factors

In blood plasma, lipids such as cholesterol and triglycer-ides are bound to various proteins (apoproteins) to form lipoproteins. The degree to which lipoproteins cause atherosclerosis depends on their type, size and concentrations in plasma. HDL do not cause atherosclerosis but on the contrary, they have antiatherogenic properties. In contrast, LDL, particularly small dense LDL, IDL (intermediate density lipoproteins), and small species of VLDL (very low density lipoproteins) are atherogenic, particularly if they are chemically modified, for example by oxidation. Chylomicrons and large VLDL are not atherogenic but high concentrations of these triglyceride-rich lipoproteins can cause pancreatitis.

11.2 Total cholesterol and LDL-cholesterol

Most of cholesterol in blood plasma is normally carried in LDL and, over a wide range of cholesterol concentrations, there is a strong and graded positive association between total as well as LDL-cholesterol and the risk of cardiovascular disease [673]. This association applies to individuals without cardiovascular disease as well as to patients with established disease. It applies to women as well as men, although the general level of risk is lower in women until the menopause, and it applies to old as well as younger people [674, 675].

This association is considerably modified by other risk factors such as age, sex, smoking, blood pressure, diabetes, and low HDL measured as HDL-cholesterol [676]. However, coronary artery disease is rare in populations with total cholesterol less than 3-4 mmol/l (∼ 115-155 mg/dl), even in the presence of other risk factors. Conversely, coronary artery disease is inevitable in untreated patients with the severest forms of familial hypercholesterolaemia, even in the absence of other risk factors. In patients with the fairly common heterozygous form of familial hypercholesterolaemia, LDL-cholesterol can be quite elevated, 7-12 mmol/l (∼270-465 mg/dl), and LDL-cholesterol is extremely elevated in the rare homozygous form, 12-20 mmol/l (465-775 mg/dl).

The results of epidemiological studies, as well as trials with angiographic or clinical end points, confirm that the reduction of LDL, measured as LDL-cholesterol, must be of prime concern in both primary and secondary prevention of atherosclerotic disease.

11.3 Triglycerides

Hypertriglyceridaemia is also associated with risk of atherosclerotic disease, but the association is not as strong as it is for hypercholesterolaemia.

Although the risk of cardiovascular disease does increase with hypertriglyceridaemia [677], the risk is associated more strongly with moderate than with very severe hypertriglyceridaemia, probably because the former is often due to accumulation in plasma of triglycericeride-rich atherogenic IDL and small VLDL, whereas the latter can be due to nonatherogenic large VLDL and chylomi-crons [678]. Triglycerides levels vary on the basis of the length of fasting (no. of hours) and particular conditions (last heavy meal, alcohol consumption, smoking etc.) before blood sampling.

The association of hypertriglyceridemia to atherosclerosis can be explained by direct atherogenic effects of IDL and small VLDL on the vessel wall, by the fact that high concentrations of triglycerides are also commonly attended by low concentrations of HDL, by the relationship of hypertriglyceridaemia and thrombogenesis as well as by the fact that hypertriglyceridemia can be associated with a large number of physiological and environmental phenomena that promote the development of early-onset cardiovascular disease. They include type 2 diabetes, hypertension, hyperinsulinemia/insulin resistance, abdominal obesity, low physical activity and low consumption of fruits and vegetables.

A triglycerides value more than 1.7 mmol/l (∼ 150 mg/dl) is considered a marker of increased risk, but concentrations less than 1.7 mmol/l are not considered a goal of therapy.

11.4 HDL

Low concentrations of HDL, measured as HDL-choles-terol, are clearly associated, not only with early development of atherosclerosis, but also with poor outcome in those who already have cardiovascular disease [679, 680]. The association is not invariable, since it is not apparent in societies in which the risk of atherosclerotic cardiovascular disease is low [681]. Therefore, it has to be stressed that smoking, sedentary lifestyle, obesity and type 2 diabetes cause lower HDL-cholesterol.

The combination of moderately elevated triglycerides and low concentrations of HDL-cholesterol is very common in patients with type 2 diabetes, abdominal obesity, insulin resistance and physical inactivity having a high risk for an early-onset atherosclerotic disease. It is part of a pattern of deranged plasma lipoproteins characterized by a triad of increased concentrations of IDL and VLDL, the presence of small dense LDL, and low concentrations of HDL. It may even rival hyperch-olesterolaemia, due to high concentrations of LDL, as a cause of atherosclerosis. However, a practical limitation is that one component of this triad, atherogenic small dense LDL, cannot be measured in routine clinical practice but low HDL cholesterol and moderate hypertriglyceridemia suffice to indicate the need for preventive treatment of these patients.

HDLs are antiatherogenic [682] and it has been suggested that HDLs exert this effect through anti-inflammatory, antithrombotic and antiapoptotic mechanisms. They also inhibit expression of adhesion molecules thus inhibiting the adhesion of monocytes to endothelial cells, an early step in the atherosclerotic process. They stimulate the efflux of cholesterol from foam cells, they stimulate prostacyclin synthesis and inhibit the synthesis of platelet-activating factor in endothelial cells which all contributes to their protective role. The participation of HDL in the transportation of cholesterol to the liver from other organs and tissues containing a surplus of cholesterol (termed reverse cholesterol transport) is yet another mechanism by which HDL could protect the artery wall.

HDL-cholesterol is not considered a goal of therapy in the present document. Instead, HDL-cholesterol less than 1 mmol/l (∼40 mg/dl) in men and less than 1.2 mmol/l (∼ 45 mg/dl) in women is considered a marker of increased risk that should suggest to the physician that attention to lifestyle and management of high LDL-cholesterol, high blood pressure, smoking and obesity is necessary.

11.5 Other lipoproteins and lipoprotein components

11.6 Calculated lipoprotein variables

11.7 Management of dyslipidaemia

11.7.4 Drugs

In most European countries, the current armamentarium of lipid-lowering drugs includes inhibitors of HMG CoA reductase (statins), fibrates, bile acid sequestrants (anion exchange resins), nicotinic acid and selective cholesterol absorption inhibitors such as ezetimibe. All of these drug classes, with the exception of cholesterol absorption inhibitors, have been shown in trials to reduce myocardial infarction and coronary death.

The most convincing evidence from clinical end point trials as well as from angiographic and intravascular ultrasound (IVUS) trials demonstrating inhibition of the progression and/or regression of atherosclerosis has been obtained with the most potent of the lipid-lowering drugs – the statins. Statins are therefore first line drugs for lowering LDL-cholesterol. There are differences among statins in terms of their LDL-lowering efficacy. Among all lipid-lowering drugs, only the most potent statins in the highest doses have been shown to halt progression or induce regression of coronary atherosclerosis as found in two IVUS studies [572, 702]. Pleiotropic effects (such as inhibition of LDL oxidation, anti-inflammatory, and antithrombotic) have been suggested based on various experimental systems; it has been claimed that these mechanisms are important for statin efficacy. However, the major mechanism of their antiatherogenic properties is attributable to their lipid-lowering effect. This is confirmed in studies with more potent statins [572, 712] or higher doses of the same statin [713] in head-to-head comparison trials, as evidenced by reduced atheroma [572] or less risk of major cardiovascular events [712, 713] with more aggressive treatment. Lipid-lowering efficacy as the main determinant of the statin effects is also supported by a recent large meta-analysis of 14 statin trials [99] showing that the reduction of major vascular events was proportional to the absolute reduction of LDL levels. No significant differences suggesting different pleiotropic effects among statins were found in another recent analysis comparing studies with atorvastatin, simvastatin and pravastatin [714]. A tendency towards lower event rates with more potent statins, however, was observed in this analysis. It should be noted that no direct head-to-head comparisons between statins and other lipid-lowering drugs have been performed in appropriate trials. Pleiotropic effects unique to statins as a class therefore cannot be excluded, although a recent meta-regression analysis comparing 10 statin trials with 8 trials using other treatments (diet, bile acid sequestrants or surgery) found no additional effect for statins beyond that of LDL-lowering [715]. Unfortunately, head-to-head comparisons on clinical end points of different statins and different dosages of the same statin are scarce.

All the large trials have confirmed the good safety of statins [99] and these drugs are easy to use. The most serious side-effects are myopathy and, very rarely, rhabdomyolysis. Because statins are prescribed on a long-term basis possible interactions with other drugs deserve particular attention, as many patients will receive pharmacological therapy for concomitant conditions during the course of statin treatment [716] (Table 11).

Fibrates lower triglycerides and increase HDL quite effectively but they lower total and LDL-cholesterol much less than the statins. They are also easy to use. Since the evidence from clinical trials to support the wide-spread use of fibrates was not as good as that supporting statins, they were considered useful only for treatment of dyslipidemic patients with low HDL, high triglycerides, and other characteristics of the insulin resistance syndrome and type 2 diabetes. In FIELD, a large randomized controlled trial with high-risk diabetic patients, fenofibrate reduced only nonfatal myocardial infarctions and revascularization, but no beneficial effect on risk of fatal coronary events was observed [717]. Considering the much more convincing evidence for the efficacy of statins in diabetic patients [696, 701], fibrate monotherapy cannot be recommended as first-line treatment in this high-risk group, but may be considered in those with persistently low HDL levels. They may also be considered in those with severely elevated triglycerides, primarily to prevent complications such as pancreatitis.

OPEN IN VIEWER

Table 11

Selected drugs that may increase risk of myopathy and rhabdomyolysis when used concomitantly with statin CYP3A4 inhibitors/substrates

•Cyclosporine, tacrolimus

•Macrolides (azithromycin, clanthromycin, erythromycin)

•Azole antifungas (itraconazole, ketoconazole)

•Calcium antagonists (mibefradil, diltiazem, verapamil)

•Nefazodone

•Protease inhibitors (amprenavir, indinavir, nelfinavir, ritonavir, saquinavir)

•Sildenafil

•Warfarin

•Others: digoxin, niacin, fibrates (gemfibrozil)

Anion-exchange resins and niacin (nicotinic acid) are also effective lipid-lowering agents. They can be difficult to use, however, and annoying side effects such as constipation and flushing, respectively, today usually limit the use of these drugs to lipid specialists. However, data suggest that nicotinic acid is more effective in increasing HDL-cholesterol than fibrates [718] and the long-term follow-up results of the coronary drug project indicated a reduction in total mortality with nicotinic acid [719].

An inhibitor of cholesterol absorption from the small intestine (ezetimibe) has become available in most European countries. In monotherapy ezetimibe has mild LDL-lowering effects and can be considered in patients with active liver disease, when statin, fibrate and nicotinic acid are contraindicated and in patients having adverse effects on statins. The main role of ezetimibe, however, is in combination therapy with statins in patients not reaching targets with statin monotherapy [720–722]. Such combination therapy has shown good LDL-lowering efficacy. No trials with clinical end points, however, have been reported.

11.8 Therapy of raised triglycerides

Some patients with raised blood triglyceride levels show a marked response to alcohol or weight reduction which form the first line of management

Bile acid sequestrants (anion-exchange resins) tend to increase triglycerides, and they should only be used when triglycerides are less than 2 mmol/l (∼ 180 mg/dl) or if given in conjunction with triglyceride-lowering agents. Statins are usually used for patients with triglycerides up to 5 mmol/l (∼ 450 mg/dl). When triglycerides are between 5 and 10 mmol/l (∼ 450-900 mg/dl), either fibrates or statins may be used as first choice drugs, and niacin is a good drug in selected patients. Fish oils are also triglyceride-lowering agents and might be useful as a third-line therapy for patients with hypertriglyceridaemia resistant to or intolerant of fibrates or niacin or in combination with other triglyceride lowering drugs. When triglycerides exceed 10 mmol/l (∼900 mg/dl), drugs are generally not useful, but fibrates may still be tried in order to prevent pancreatitis. If ineffective, triglycerides must be reduced by restriction of alcohol, treatment of diabetes with insulin, withdrawal of estrogen therapy, etc. In the rare patients with severe primary hypertriglycer-idaemia, it is necessary to restrict absolutely the intake of alcohol and severely restrict long-chain fat of both animal and vegetable origin.

11.9 Drug combinations: effects and side-effects

Lipid-lowering drugs can be used in combination and in some patients this is necessary to achieve the treatment goals both in familial hypercholesterolaemia and in combined hyperlipidaemia. In familial hypercholestero-laemia, for example, the combination of a bile-acid sequestrant and a statin is very useful, although a combination of a statin with cholesterol absorption inhibitor ezetimibe appears a more potent and better tolerated option [720–722]. A statin and niacin can be used in some patients. Statins can also be combined with fibrates, but this combination has been associated with higher incidence of myopathy, even fatal rhabdomyolysis, and patients must be carefully selected and carefully instructed about warning symptoms (myalgia). However, these adverse effects are very rare and should not be the cause to deny the combined treatment to patients who really need it. The combination of fibrate with ezetimibe may be an option in selected patients with mixed hyperlipidaemia. Although efficacy and safety of such treatment was good in one study [723], other studies and long-term data are awaited.

11.10 Lipid-lowering therapy in acute coronary syndrome

The MIRACL trial in patients with acute coronary syndromes showed that treatment with a statin, initiated within 4 days, can reduce the recurrence of myocardial ischaemia during the following 4 months [724]. The same has been shown in the PROVE-IT trial [725], although the A-to-Z trial did not show any significant benefit over the first 4 months [726]. Therefore statins should be initiated while patients are in hospital with an acute coronary event. Another strong reason to initiate statin therapy while patients are in hospital is the fact that adherence rates to statins have been shown to be much higher if patients are discharged with a prescription. Plasma lipids should be reevaluated both at 4-6 weeks and 3 months after acute event and/or initiation of the lipid-lowering therapy to evaluate whether target levels have been achieved and to screen for liver dysfunction. The second control at 3 months is important for patients after myocardial infarction, since decrease of plasma lipids due to acute phase reaction may last for such periods of time. The early drug treatment should nevertheless be combined with effective lifestyle changes and particularly dietary intervention after the hospital discharge. The strategy to ensure that goal concentrations are reached must obviously depend on the organization of medical care in each European country.

11.11 LDL apheresis

Rare patients with severe hypercholesterolaemia, especially homozygous familial hypercholesterolaemia, require specialist evaluation of the need for LDL apheresis. By this demanding and expensive but effective technique, LDL is removed from plasma during extracorporeal circulation weekly or every other week. LDL apheresis has to be combined with treatment with statins.

11.12 Goals of therapy

Physiological concentrations of LDL-cholesterol are probably around 1-2 mmol/l (∼ 40-80 mg/dl), but whether clinical benefit results from reducing LDL-cholesterol to such low levels has been the subject of some controversy. Several studies such as 4S [467] and LIPID trials [694] indicated that there is no lower threshold for benefits of total and LDL-cholesterol level reduction. The results of the Heart Protection Study [63] demonstrated the same degree of benefit, given in relative terms, of lowering LDL-cholesterol from 3 to 2 mmol/l as from 4 to 3 mmol/l, suggesting again that, at least down to about 2 mmol/l, there is no clear threshold value for benefit. This conclusion is also supported by the result of the ASCOT-LLA trial, which demonstrated clinical benefit from reducing LDL-cholesterol by about 1.2 mmol/l from a baseline level of only 3.4 mmol/l [469]. It is also consistent with observational epidemiology [725]. In patients with cardiovascular disease, the Treat to New Targets (TNT) study showed that an extra benefit could be obtained by larger reductions of LDL-cholesterol without increasing the risk of serious adverse events [713]. The IDEAL study had similar findings with a composite end point that included stroke [727].

In 1998 and 2003, the Joint European Societies recommended that reduction of total cholesterol below 5 mmol/l (∼ 190 mg/dl) and LDL-cholesterol below 3 mmol/l (∼ 115 mg/dl) could be goals of therapy consistent with the evidence available at that time [2, 3]. However, for patients with established cardiovascular disease and patients with diabetes the treatment goals were lower: total cholesterol less than 4.5 mmol/l (∼ 175 mg/dl) and LDL-cholesterol less than 2.5 (∼ 100 mg/dl).

The results of clinical trials since then raise the question of whether the goals for total cholesterol and LDL-cholesterol be lowered further. It is still not clear what the ideal LDL-cholesterol value is, but there is evidence of benefit down to 2 mmol/l and even lower in all patients with established atherosclerotic disease. In several studies benefit was seen with more intensive therapy reaching even lower levels (1.6-1.8 mmol/l) and in one noncontrolled study (ASTEROID) mean LDL of 1.55 mmol/l was associated with regression of atheroma [702]. Therefore an optional target for high risk patients of less than 2.0 mmol/l has been defined if feasible clinically and in terms of local financial resources.

Asymptomatic people at high risk of developing CVD, whose untreated values of total and LDL-cholesterol are already close to 5 (∼ 190 mg/dl) and 3 mmol/l (∼ 115 mg/dl), respectively, will definitely benefit from further reduction of total cholesterol to less than 4.5 mmol/l (∼ 175 mg/dl) and from further reduction of LDL-cholesterol to less than 2.5 mmol/l (∼ 100 mg/dl) with lipid-lowering treatment.

However, these goals cannot be reached with the same ease by all patients. Patients with concentrations of plasma lipids that are only slightly abnormal can reach these goals of therapy fairly easily with diet and moderate doses of drugs. When these goals have not been reached in asymptomatic people at high risk they will still benefit to the extent that cholesterol has been lowered.

11.13 Lipids and stroke risk

Cholesterol has a different history as a risk factor for cerebrovascular disease than for CHD. Only few early epidemiological studies have found increasing risk of stroke with high total cholesterol [728], and early meta-analyses have not found any significant association between stroke and total cholesterol levels [703]. However, this was due to the fact that in these studies ischaemic stroke was considered together with haemorrhagic stroke. When ischaemic stroke alone was analysed, a clear correlation with total and LDL-cholesterol levels was found indicating that high cholesterol is a risk factor for ischaemic stroke, but not for haemorrhagic stroke [729]. Conversely, an increased risk of haemorrhagic stroke has been documented in patients with extremely low cholesterol levels. Increased level of triglycerides was also significantly associated with nonhaemorrhagic strokes [730], as was decreased HDL-cholesterol [731]. Almost all big multicentre randomized controlled trials with statins have shown a significant reduction in cerebrovas-cular events including nonhaemorrhagic strokes in patients treated with statins. This was true for studies with simvastatin [63, 467], atorvastatin [469] and several trials with pravastatin [692, 694]. However, in one trial with pravastatin in elderly, despite a significant reduction in transient ischaemic attacks (TIAs), no significant reduction in strokes could be proven [695]. Many stroke patients have CHD and/or diabetes and they will clearly benefit from statins [63]. Intensive cholesterol lowering with 80 mg atorvastatin/day in a randomized, placebo-controlled trial of patients with unstable angina or non-Q-wave myocardial infarction reduced after only 16 weeks the overall stroke rate by half without increasing the risk of haemorrhagic stroke [732]. A recently published trial evaluated the effects of 80 mg atorvastatin/day in patients with ischaemic cerebrovascular disease. A significant risk reduction for recurrent cerebrovascular events has been shown but patients receiving atorvastatin had some more haemorrhagic strokes [733]. Similar beneficial effects of fibrates on risk reduction could not be proven neither for overall nor for nonhaemorrhagic stroke [734].

12.1 Hyperglycaemia, diabetes and CVD risk

Epidemiological studies demonstrate a linear association between increasing glucose levels and the risk of developing CVD continuing all the way down to the normal range. This has been demonstrated both using the 2-h value following an oral glucose tolerance test [735] and using the integrated measure of glycated haemoglobin HbA1c [736–741]. In the nondiabetic range, nonfasting plasma glucose values are more predictive than fasting values in relation to CVD, but these studies have compared fasting plasma glucose values with the 2-h value following an oral glucose tolerance test. The clinically more relevant comparison would be between fasting and postprandial glucose, but this study has not been performed.

Individuals can be classified into different categories (Table 13) based on fasting plasma glucose and 2-h plasma glucose measurement following a 75 g oral glucose tolerance test [742, 743].

OPEN IN VIEWER

Table 13

Classification based on plasma glucose levels

	Fasting plasma glucose	2h plasma glucose
Normal	≤ 6.0 mmol/l (≤ 108 mg/dl)	>7.8 mmol/l (<140 mg/dl)
Impaired Fasting Glycaemia (IFG)a	6.1 to 6.9 mmol/l (110 to 124 mg/dl)	>7.8 mmol/l (<140 mg/dl)
Impaired Glucose Tolerance (IGT)	>7.0 mmol/l (<126 mg/dl)	7.8 to 11.0 mmol/l (140 to 198 mg/dl)
Diabetes	≥ 7.0 mmol/l (≥126 mg/dl)	≥ 11.1 mmol/l (≥200 mg/dl)

^aThe American Diabetes Association use ≥ 5.6 mmol/l (≥ 100 mg/dl) as the lower cut-point for IFG. This lower cut-point was not adopted in Europe, and is not recommended by WHO [743].

12.2 Hyperglycaemia and CVD risk

Epidemiological studies have shown that hyperglycaemia is associated with an increased risk of developing CHD as well as other atherosclerotic diseases [736–740]. This is true for diabetes as well as for individuals with impaired glucose tolerance (IGT) – the intermediate stage between normal and diabetic postload glucose levels. In diabetic individuals the relative risk of CVD is in the order of 2-4, while in individuals with IGT the relative risk is 1.5 compared to individuals with normal glucose tolerance [744]. Isolated IFG (fasting plasma glucose in the IFG range and normal 2 h plasma glucose) is not associated with an increased risk of CVD.

'Diabetes’ is classified into two major groups (type 1 and type 2) and a number of smaller specific entities classified on the basis of specific genetic markers, syndromes or secondary diabetes induced by other diseases, chemicals or drugs [742].

Type 1 diabetes is characterized by loss of β-cell function and endogenous insulin production to a level where the individual would die from ketoacidosis if not treated with insulin. The incidence is highest in children and young adults, but type 1 diabetes can develop at any age.

Type 2 diabetes is a condition characterized by a combination of insulin resistance and β-cell failure. With increasing duration, insulin may be necessary in up to 50% of the patients to obtain acceptable metabolic control [745]. The incidence and prevalence increases by age, but the condition is heavily associated with obesity and lack of physical activity, and for this reason the incidence is not only increasing world wide [746], but the age at diabetes onset is also decreasing, and type 2 diabetes can be seen even in teenagers. An important underlying mechanism leading to type 2 diabetes is insulin resistance, which again is associated with a long list of cardiovascular risk factors including hypertension, dyslipidaemia, endothelial dysfunction and microalbuminuria [747]. This clustering of risk factors partly explains the increased risk of CVD associated with diabetes and glucose intolerance. However, many studies have reported an association independent of traditional and nontraditional risk factors.

12.3 Risk of CVD, CHD and stroke in diabetes

The association between these macrovascular complications differs markedly between type 1 and type 2 diabetes. In type 1 diabetic patients there is a two to three-fold increase in the risk of developing CVD, CHD and stroke. This increased risk is almost entirely confined to the patients developing diabetic renal disease [748].

In type 2 diabetes, all patients are at increased CVD risk, even in the absence of diabetic nephropathy. Finnish data published in 1998 suggested that the risk of developing a myocardial infarction in patients with type 2 diabetes is of the same order as for patients without diabetes who have already suffered their first MI [749]. This had immense effect on treatment guidelines, where diabetes was labeled as a ‘CVD-equivalent’ when it came to risk assessment. Since then, however, many studies based on different study cohorts have addressed this issue and it has become clear that this is an oversimplification and that the impact of type 2 diabetes on CVD risk is influenced by a number of factors, including the duration of diabetes, age and sex [290, 750–760]. The effect of gender on the diabetes-related CVD is of major importance in the interpretation of study results. Because the relative impact of type 2 diabetes on the CVD risk is much stronger in women than in men, the concept of diabetes as CVD-equivalent has proved to be true in women in studies analysing data separately for both sexes [751, 754, 756–760]. In men the outcomes of studies have been less consistent, but in some study cohorts type 2 diabetes has been CVD-equivalent also in men, particularly in older age groups [756, 758–760].

Although a substantial proportion of the excess risk of atherosclerotic disease in both type 1 and type 2 diabetes is caused by the diabetic state itself and related factors, from the point of prevention of atherosclerotic disease it is important to emphasize that the conventional, modifiable major cardiovascular risk factors, elevated blood pressure, elevated LDL-cholesterol, and smoking show in both type 1 [761], and type 2 diabetic patients [762] similar relationships with the risk of CVD as in nondiabetic patients. Because diabetes itself increases the absolute risk of cardiovascular disease, the additional impact of conventional risk factors leads to a more dramatic increase in absolute risk than in nondiabetic patients and thus the modification of these risk factors offers a great potential for prevention. Consequently, individualized global risk assessment and individualized prevention strategies are even more important in individuals with diabetes than in nondiabetic patients.

12.4 The evidence for the current recommendations on prevention of CVD in diabetes

With the exception of glucose management, prevention of CVD follows the same general principles as for people without diabetes. A multifactorial approach to treatment and achieving low BP and low LDL are particularly important, thus many of the treatment targets are tougher for patients with diabetes.

12.5 Metabolic syndrome

The name ‘metabolic syndrome’ has been given to a clustering of several cardiovascular risk factors (obesity and its central distribution, elevated plasma glucose, elevated plasma triglycerides, low HDL cholesterol and raised blood pressure). The pathogenesis of the metabolic syndrome is complex and incompletely understood, but insulin resistance and accumulation of intra-abdominal fat tissue are considered to be important underlying factors. There is still an ongoing debate regarding whether this clustering of risk factors really represents an entity in its own right and whether it predicts the risk of cardiovascular disease more strongly than its individual components. The metabolic syndrome has, however, become a subject of great interest because of its association with the risk of development of type 2 diabetes and atherosclerotic CVD.

Several international or national expert groups have during the last years formulated definitions for the metabolic syndrome. The first of them, given by the WHO consultation report in 1999 [778] was primarily intended for research purposes (Table 14). Various modifications of this definition have been widely used in epidemiological research, usually substituting fasting plasma insulin above the highest quartile for insulin resistance (as assessed by hyperinsulinaemic euglycaemic clamp glucose uptake below lowest quartile) and omitting microalbuminuria, as proposed by the European Group for the Study of Insulin Resistance (EGIR) [779]. The definition given by the National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III) in 2001 [689] was created for clinical use and has become widely adopted. A revised version of the NCEP-ATP III definition was recommended in 2004 by the American Heart Association (AHA) and the National Heart, Lung, and Blood Institute (NHLBI) [780] (Table 15). The newest definition was launched in 2005 by the expert group of the International Diabetes Federation [781] (Table 16). The major components included in these definitions of the metabolic syndrome are more or less similar but there are differences in the emphasis between them and in the diagnostic threshold values given for individual components. The WHO definition includes impaired glucose regulation and/or insulin resistance as a prerequisite. The NCEP-ATP III definition does not have any prerequisite and aims at detection of increased CVD risk. The IDF definition has the presence of central obesity as its starting point.

OPEN IN VIEWER

Table 14

The WHO definition of the metabolic syndrome

Impaired glucose regulation (IFG, IGT or type 2 diabetes) and/or insulin resistance (in hyperinsulinaemic, euglycaemic clamp glucose uptake below lowest quartile for background population)

and at least two of the following four components:

Central obesity: waist hip ratio >0.90 in men, >0.85 in women and/or BMI >30 kg/rn2

Dyslipidaemia: plasma triglycerides ≥1.7 mmol/l (≥150 mg/dl) and/or plasma HDL-cholesterol >0.9 mmol/l (>35 mg/dl) in men, >1.0 mmol/l (<39 mg/dl) in women

Raised blood pressure: ≥140/90 mmHg

Microalbuminuria: urinary albumin excretion rate ≥20 μg/min or albumin:creatinine ratio ≥ 30 mg/g

OPEN IN VIEWER

Table 15

Original and revised NCEP-ATP III definitions of the metabolic syndrome

At least three of the following five components:

Central obesity: waist circumference>102 cm in men, >88 cm in women

Elevated triglycerides: ≥ 1.7 mmol/l (≥ 150 mg/dl)

Low HDL cholesterol: >1.03 mmol/l (>40 mg/dl) in men, >1.29 mmol/l (<50 mg/dl) in women

Raised blood pressure: SBP ≥130 mmHg and/or DBP ≥85 mmHg or treatment of previously diagnosed hypertension

Impaired fasting glycaemia: fasting plasma glucose ≥6.1 mmol/l (≥110 mg/dl) [≥ 5.6 mmol/l (≥ 100 mg/dl)]a or previously diagnosed type 2 diabetes

^aThe revised version recommended by the AHA/NHLBI uses the lower cut-off for impaired fasting glycaemia

OPEN IN VIEWER

Table 16

The International Diabetes Federation definition of the metabolic syndrome

Central obesity, defined by ethnic-specific waist circumference criteria (≥ 94 cm for Europid men, ≥ 80 cm for Europid women)

and any two of the following four components:

Elevated triglycerides: ≥ 1.7 mmol/l (≥ 150 mg/dl) or specific treatment for this lipid abnormality

Low HDL cholesterol: >1.03 mmol/l (>40 mg/dl) in men, >1.29 mmol/l (< 50 mg/dl) in women or specific treatment for this lipid abnormality

Raised blood pressure: SBP ≥ 130 mmHg and/or DBP ≥85 mmHg or treatment of previously diagnosed hypertension

Impaired fasting glycaemia: fasting plasma glucose ≥ 5.6 mmol/l (≥ 100 mg/dl) or previously diagnosed type 2 diabetes

In the revised NCEP-ATP III definition, recommended by the AHA/NHLBI, the cut-off for impaired fasting glycaemia was lowered to ≥ 5.6 mmol/l (≥ 100 mg/dl) following recommendations given by the ADA. If fasting plasma glucose is ≥ 5.6 mmol/l (≥ 100 mg/dl), oral glucose tolerance testing is strongly recommended, but is not necessary to define the presence of the syndrome.

The majority of patients with type 2 diabetes have the risk factor characteristics of the metabolic syndrome and the presence of this risk factor clustering has an adverse effect on their prognosis. The diagnosis of the metabolic syndrome is, however, of greatest importance in nondiabetic patients as an indicator of an increased risk of developing type 2 diabetes and CVD.

In the DECODE Study population, pooling data from nine European population based cohorts of men and women aged 30-89 years, the overall prevalence of the metabolic syndrome was in nondiabetic men with a modified WHO definition 25%, with the original NCEP-ATP III definition 23%, with the revised NCEP-ATP III definition 30% and with the IDF definition 34%. In nondiabetic women the overall prevalence of the metabolic syndrome with these definitions was 18, 21, 26 and 32%, respectively [782]. As can be expected because of differences between the definitions of the metabolic syndrome, their concordance in the identification of subjects with the syndrome is not good. Of all patients fulfilling the criteria for any one of these definitions, only about one third fulfilled the criteria for all the definitions.

Some information is available from prospective cohort studies on the predictive value of different definitions of the metabolic syndrome with regard to the development of type 2 diabetes. A 4-year follow-up study of Finnish middle-aged men showed that men with the metabolic syndrome either with the WHO definition or the NCEP-ATP III definition were at high risk of developing diabetes; the odds ratios were from 5.0 to 8.8, depending on the cut-offs used for the measures of central obesity [783]. In a systematic review and meta-analysis of studies on the relationship of the metabolic syndrome with the original NCEP-ATP III definition to the development of type 2 diabetes the combined relative risk was 2.99 [784].

The relationship of the metabolic syndrome with the risk of CVD has been assessed in many prospective cohort studies, using mainly some modification of the WHO definition or the original NCEP-ATP III definition. A systematic review and meta-analysis of such studies showed that for studies that used the most exact WHO definition the combined estimate of relative risk for CVD was 1.93 and 2.60 for CHD [784]. For studies using the original NCEP-ATP III definition the combined estimate of relative risk was 1.65 for CVD. In the DECODE Study population, combining data from nine European cohorts with a maximum follow-up of 7 to 16 years, the hazard ratios of different definitions of the metabolic syndrome for CVD mortality, adjusting for age, serum cholesterol and smoking, were as follows in men: with the modified WHO definition 2.09, with the original NCEP-ATP III definition 1.74, with the revised NCEP-ATP III definition 1.72, and with the IDF definition 1.51. In women the hazard ratios were somewhat lower: 1.60, 1.39, 1.09, and 1.53, respectively. Some other studies using either the original NCEP-ATP III definition or the modified WHO definition have, however, shown similar or even higher hazard ratios for CVD in women with the metabolic syndrome as compared with men with the syndrome [322, 785–787].

An important issue is whether identification of patients with the metabolic syndrome will bring important additional information about CVD risk over and above that obtained from general multifactorial CVD assessment tools, such as the Framingham or SCORE risk equations based on European cohort study data. This question was addressed in a study based on the follow-up data of the nondiabetic men aged 50-69 years in the DECODE Study population [788]. Of these men 51% had an estimated 10-year risk of CVD death under 5%, using the SCORE risk equation, and 22% of them had the metabolic syndrome with the NCEP-ATP III definition. In these low risk men, the hazard ratio for fatal CVD was 2.71 for men with the syndrome compared to men without the syndrome. A large waist circumference (> 102 cm) carried a hazard ratio of 2.24 in low risk men. In men with a 10-year risk of CVD death ≥ 5%, the presence of the metabolic syndrome did not add substantially to the risk predicted by the SCORE equation. These findings indicate that the diagnosis of the metabolic syndrome may identify patients with increased risk of CVD among those who would become classified as low-risk individuals using conventional tools for CVD risk assessment. A large waist circumference appears to be a useful warning sign and should stimulate assessment of other components of the metabolic syndrome.

12.6 Management of risk in clinical practice

12.6.3 Blood pressure

Targets for blood pressure are generally more ambitious in patients with diabetes. This is first of all due to the fact that very strict control of blood pressure is the most important single factor preventing development of diabetic nephropathy and end stage renal failure [459–461, 573, 804]. Furthermore, subgroup analyses of the diabetic patients in trials focused on prevention of CVD in patients with hypertension have demonstrated more beneficial treatment effects in the diabetic group in the nondiabetic group, and suggests that stricter treatment targets are indicated [470, 768, 805]. The HOPE study showed that in patients with initial blood pressure of 130/79 mmHg a further, small reduction of BP (3/1 mmHg) was associated with a further reduction in cardiovascular risk [573]. The optimal BP levels to be achieved cannot be precisely defined, but values below 130/80 may be desirable in diabetic patients. In diabetic patients with diabetic nephropathy and proteinuria >lg/24h, values as low as 125/75 mmHg or lower are recommended if achievable without unacceptable side effects.

The type of antihypertensive medication also seems to be important. Angiotensin converting enzyme (ACE) inhibitors and angiotensin II receptor inhibitors have proven to be particularly effective in preventing progression from microalbuminuria to overt nephropathy in type 1 as well as in type 2 diabetic patients [459–461, 573, 804]. Thus in these groups of patients, ACE inhibitors and angiotensin II receptor blockers would be preferred as initial therapy; however, most patients will require a combination of two or more drugs [666]. Combination therapy including both an ACE inhibitor and an angiotensin II receptor antagonist has been shown to have additional beneficial effect, over and above the effect of each on progression of diabetic renal disease, and consequently this ‘dual blockade’ principle may be beneficial also in preventing CVD.

OPEN IN VIEWER

Table 17

Glucose control assessment levels in patients with type 2 diabetes as recommended by different organizations

Organization	HbA1C (%)	Fasting plasma glucose (mmol/l)	Post-prandial plasma glucose (mmol/l)
ADA	>7.0	5.0-7.2 (90-130 mg/dl)	< 10.0 (180 mg/dl)
AACE	≤ 6.5	≤ 7.0 (≤ 110 mg/dl)	≤ 7.8 (≤ 140 mg/dl)
IDF-Europe	≤ 6.5	≤ 6.0 (≤ 108 mg/dl)	≤ 7.5 (≤ 135 mg/dl)
IDF-Global	>6.5	>6.0 (>110 mg/dl	>8.0 (<145 mg/dl)

ADA, American Diabetes Association; AACE, American Association for Clinical Endocrinologists; IDF, International Diabetes Federation.

In diabetic patients with hypertension and established coronary heart disease, particularly those who have survived a myocardial infarction and in those with angina pectoris, the use of β-blockers is indicated.

12.7 Precursors of diabetes

Impaired glucose tolerance (IGT) and impaired fasting glycaemia (IFG) are both conditions associated with an increased risk of developing type 2 diabetes, and IGT is associated with a deterioration of the cardiovascular risk profile [806] and increased risk of death from all causes as well as from CVD, CHD and stroke [735]. In patients with IGT, several studies have demonstrated that progression to diabetes can be prevented or delayed by lifestyle intervention, and thus these patients should be identified where possible and provided with necessary support.

Increasing levels of metabolic risk factors are seen across the spectrum of nondiabetic glucose values [807], and hyperinsulinaemia is associated with all the components of the metabolic syndrome. Individuals with the metabolic syndrome are at high risk of developing cardiovascular disease, and in these individuals a total risk assessment based on the existing risk engines should be performed to assess risk, and to identify the most important risk factors available for intervention.

12.8 Prevention in patients with the metabolic syndrome

The diagnosis of the metabolic syndrome is of greatest importance in nondiabetic patients as an indicator of an increased risk of developing type 2 diabetes and CVD. It is, however, important to emphasize that interest in the metabolic syndrome should not displace the use of conventional CVD risk assessment tools, such as SCORE and other similar risk scoring tools, from their primary place in the identification of individuals who are at high CVD risk. In fact, the components of the metabolic syndrome, with the exception of the measures of central obesity, triglycerides, IFG and IGT, are included among the risk factor measurements used in conventional risk assessment systems. Adding waist circumference measurement to this set will give possibilities to detect the presence of the metabolic syndrome and help identify people who actually are at high risk of CVD, although they do not get particularly high risk scores in conventional CVD risk assessment. The original and revised NCEP-ATP III definitions and the IDF definition of the metabolic syndrome are suitable for clinical use, but it is important to realise that because of a lowered threshold for IFG in the revised NCEP-ATP III and IDF definitions and a lowered threshold for central obesity in the IDF definition, these definitions will pick up a larger proportion of people and will have a lower positive predictive value than the original NCEP-ATP III definition.

Since lifestyles have a strong influence on all the components of the metabolic syndrome, the main emphasis in the management of the risk in people with this syndrome should be in professionally supervised lifestyle change, particularly directed to the reduction of overweight and increased physical activity. Although the dyslipidaemia of the metabolic syndrome is characterized by elevated triglycerides and/or low HDL cholesterol, lipid management should, however, be steered with LDL-cholesterol goals in mind. Subgroup analyses of large statin trials have shown that coronary heart disease patients with and without the metabolic syndrome get from statin treatment a similar substantial relative reduction of CVD events, but the absolute benefit may be even greater in those with the syndrome, because they are at higher absolute risk [808, 809].

13.1 Low socio-economic status

Men and women with low socio-economic status (SES), defined as low education, holding a low status job or living in a poor residential area, have an increased all-cause as well as CHD mortality risk, which is only in part mediated by traditional risk factors [811]. Low SES persons are also more likely to experience low control at work, social isolation and lack of social support, poor capacity to cope with stressors, hopelessness and depressed mood. These mediators can be modified and preventive efforts should be directed towards those of low SES, because they are in greater need. In turn, leaving low SES unrecognized when applying standardized risk charts may underestimate CVD risk in socially disadvantaged persons [817], lead to insufficient preventive efforts and further increase CVD burden in this disadvantaged group.

13.2 Social isolation and low social support

Social support has beneficial effects on lifestyle and health behaviour. People who are isolated or disconnected from others are at increased risk of dying prematurely from ischaemic heart disease [818–820]. Similarly lack of social support leads to decreased survival and poorer prognosis among people with clinical manifestations of CVD [821].

13.3 Psychosocial stress at work and outside work

Stress at work, characterized by combinations of high demand with low control or high effort with low reward, predicts CHD risk [822]. Although these effects are seen in both women and men [823, 824], work stress appears to be more relevant for cardiovascular end points in men than in women [825].

Prolonged exposure to work at irregular hours, including work at night, increases CHD risk, with a higher risk as the number of years in shift work increases [826–828]. A direct and causal effect of shift work on CHD is suggested in women [829] and in men [830].

Conflicts, crises and long-term stressful conditions in family life have also been shown to increase CHD risk, especially in women [825, 831].

13.4 Hostility

Hostility has been identified as a risk factor for CVD [832–836] and for virtually any physical illness, although effect sizes appear to be small [837].

13.5 Depression

Clinical depression, depressive symptoms and other negative emotions have been shown to predict incident CHD [838, 839], and worsen its prognosis [840, 841], independently of standard risk factors. For example, clinical depression is associated with at least doubled risk for major cardiac events. Further evidence suggests that the prevalence of depressive symptoms is higher in women compared to men. Increased risk is mediated by a variety of behavioural and psychophysiological mechanisms. Part of it may also be due to adverse effects mainly of tricyclic antidepressants [842, 843].

In patients with manifest coronary disease, depression has well-documented effects on cardiac symptoms, overall quality of life and illness behaviour (including increased healthcare utilization, low adherence with behaviour change recommendations or cardiac medications, and low rates of work resumption) [844]. Perceived social support seems to counteract the adverse effect of depression [845], whereas lack of support was found to reinforce the effects [846]. Prognostic risk is highest in persons with a combination of chronic negative affectivity and social inhibition [847, 848].

13.6 Clustering of psychosocial risk factors and biobehavioural mechanisms

CVD's multifactorial aetiology and the multiplicative effect of coronary risk factors mean that in attempting to reduce risk we must deal with the whole person and not with isolated risk factors. This principle is no less true when it comes to psychosocial risk factors and the biobehavioural mechanisms which mediate effects on pathogenesis and prognosis. It is now evident that psychosocial risk factors do not occur in isolation from one another, but tend to cluster in the same individuals and groups. For example, both women and men of lower SES are more likely to be depressed, hostile, socially isolated, and to engage in risky health behaviours [849–851]. In addition to risky health behaviours like smoking, high alcohol consumption, and unhealthy nutrition, persons with psychosocial risk factors such as depression are also more likely to express biological characteristics that are involved in promoting cardiovascular disease. Alterations have been found in autonomic function (including reduced heart rate variability), in the hypothalamic pituitary axis (HPA) and other endocrine markers, which affect haemostatic and inflammatory processes, endothelial function, and myocardial perfusion [810].

The tendency of psychosocial risk factors and biobeha-vioural mechanisms to cluster in the same individuals and groups has important implications for strategies to modify risk and improve quality of life.

As persons with high levels of negative affect are more likely to smoke, partly due to an antidepressant-like effect of cigarette smoke [852] attempts to help them quit smoking might be more successful if they include elements designed to reduce hostility and depression. Similarly, attempts to reduce the CVD risk in the socially disadvantaged, might be more effective by incorporating training in skills that will reduce negative feelings and increase access to positive, supportive social ties as well as self-efficacy [853]. In other words, behavioural interventions that reduce levels of psychosocial risk factors are likely to have broad benefits in terms of enabling people to be more successful in modifying unhealthy lifestyles and reducing biological consequences of stress – for example altered haemostatic and inflammatory functions – that are directly involved in pathogenesis.

13.7 Management of psychosocial risk factors

15.1 Family history

The importance of a family history of CHD as a coronary risk factor has been established by a number of studies, and is included as a risk factor in some risk prediction algorithms such as PROCAM [896]. The size of the risk associated with a family history of early CHD (usually defined as evidence of CVD or CHD in a first degree male relative >55 years and female relative >65 years) is in the range 1.5-1.7, and is independent of classical CHD risk factors including blood lipids, blood pressure, body mass index etc. [897, 898]. Although some studies find that risk is higher if found in a female relative, not all studies report this [899]. Based on this, a detailed family history of CHD, or other atherosclerotic disease should be part of the assessment of all patients with CHD and in the identification of high-risk individuals.

The risk of CHD increases (i) when an individual is closely related to a family member who has developed CHD. A history of CHD in a first degree relative (parents, brother or sister, or son or daughter) is more important than a similar history in a second degree relative (grandparent, aunt, uncle) or in a third degree relative (cousin); (ii) as the percentage of family members with CHD increases; and (iii) the younger the age at which family members develop CHD. For example, one study has found that, among women with a low Framingham risk score, carotid thickening was greater in those with a family history of CHD [900].

15.2 Phenotypes

We now know that the pathophysiology of CHD is characterized by a combination of acute events, such as plaque rupture, thrombosis and vasoconstriction, acting on a substrate of chronic processes, such as dyslipidaemia, hypertension, endothelial dysfunction, diabetes, cardiac and vascular hypertrophy and atherosclerosis. The study of the genetic determinants of all of these ‘phenotypes’ is therefore likely to be clinically relevant, and each will have their own genetic and environmental determinants.

For many of these phenotypes (measurable traits) there is good evidence for a relatively strong genetic contribution to the determination of levels, which is usually estimated by ‘heritability'. For apoproteins and lipid traits heritability varies between 40-60%, [901] meaning that genetic factors are determining around half of the between-individual differences and environmental factors the remainder. While there is less evidence for the heritability of exciting emerging risk factors such as intercellular adhesion molecule (ICAM), interleukin 6 (IL6), phospholipase A2 (PLA2) etc., the data available also suggests moderate to high heritability [902, 903]. For one established risk factor, plasma Lp(a), heritability is reported to be >90%, [904] and interestingly, variability at the locus coding for the apo(a) gene itself accounts almost all of the variance of plasma Lp(a) in normal populations [905]. Since meta-analyses show that levels of Lp(a) in the top tertile are associated with a 1.6-fold greater risk of CHD, [683] an effect which is of similar magnitude as smoking, the (a) gene would appear to be a major genetic factor for CHD.

15.3 Genotypes

A gene may predispose to CHD if it occurs in functionally different forms. With the publication of the entire human genome sequence, showing that there are 25 000 different genes, and the development of the Human Genetic Diversity database, the tools are now available to explore the impact of these variations on human disease. Functional polymorphisms are relatively common and may have neutral, beneficial or detrimental consequences if they affect regulatory or coding regions of genes. Genetic polymorphisms are defined as sequence variants that occur at a frequency greater than 1%. These include single nucleotide polymorphisms (SNPs) as well as insertion/deletion and copy number variants which have recently been reported to be much more common than previously appreciated [906]. SNPs have been used with increasing momentum to study the genetic determinants of complex diseases such as CHD, in case-control analyses and association and linkage disequilibrium studies with intermediate traits. The important issue is whether and under what circumstances such genetic information will be useful for diagnosis and patient management.

In general, the levels of CHD risk traits are influenced by both environmental and genetic factors with, in most individuals, gene variants of small or modest impact being involved. Thus an individual with, for example, high plasma cholesterol may have inherited several ‘raising’ alleles acting in combination, or they may have few such alleles but eat a cholesterol-raising diet, but most likely they have a combination of both influences. Thus, the identification of the complete list of the genetic variants that an individual has inherited should be of diagnostic or prognostic value, but how this genotype interacts with the environment of the individual will also need to be taken into account. This concept of Gene-Environment interaction is key to being able to use genetic information for accurate risk estimation [907]. In practice it means that a particular genotype will predispose to elevated CHD trait levels and thus CHD risk only in a certain environment (for example in smokers or hypertensive individuals, or those with diabetes). Understanding such interactions is likely to be of major research importance in future years, since these interactions shed light on pathophysiological processes.

A large number of ‘candidate’ genes have already been investigated in relation to CHD traits and to risk of CHD itself, and a comprehensive list is beyond the scope of this report. One of the problems in the field is that many initially exciting findings of large risk associated with a particular SNP have failed to be confirmed in later studies, [908] or at least the later reported effects are much smaller than originally found (the so-called ‘winners curse'). To overcome this meta-analyses have been used to obtain a statistically robust estimate of these effects, and several variants in genes involved in lipid metabolism [e.g. apolipoprotein E (apoE), apolipo-protein B (apoB), lipoprotein lipase (LPL), cholesterol ester transport proteins (CETP)], coagulation [plasminogen activator inhibitor 1 (PAI1), glycoprotein IIbIIIa (GIIbIIIa), factor V (FV)] and different aspects of endothelial function [endothelial nitric oxide synthase (eNOS), methylenetetrahydrofolate reductase (MTHFR), angiotensin converting enzyme (ACE)] [909] appear to be associated with statistically significant although rather modest effects on risk. For example, for the common ApoE protein variants (E2, E3 and E4), there is a strong and consistent impact on plasma lipid levels (E2 lowering and E4 raising), which translates into a modest E2 lower and E4 higher impact on CHD risk such that this genotype may explain 5-8% of the attributable risk of CHD in the population [910].

The ACE polymorphism has probably been the most extensively studied polymorphism so far, in relation to preclinical phenotypes and cardiovascular end points. One important feature of this polymorphism is that it appears to be a response modulator to a wide range of inducing factors. For example, it has been reported to modify the hypertrophic response of the heart to physical training, [911] the restenotic process after stent angioplasty, [912] the evolution of cardiac function after myocardial infarction [913] and the survival of patients with congestive heart failure, [914] and with the development of diabetic nephropathy and retinopathy [915] and with the risk of CHD in smokers [916]. Other candidate gene polymorphisms may also have the characteristic of being response modifiers to a number of stimuli.

CETP and alcohol dehydrogenase genotypes modify the relationship between alcohol consumption and plasma HDL-cholesterol [917, 918] and an amino acid variant that causes enzyme instability in the MTHFR protein affects the relationship between folate intake and plasma homocysteine [230]. These interactions also need to be more widely replicated in larger studies but if confirmed they offer potential prospects for CHD prevention through the identification of responders to deleterious factors or beneficial ones (drugs for example) by genotyping appropriate candidate genes.

15.4 DNA-based tests for risk prediction

Currently available CHD risk prediction algorithms [896, 919] have a very low prediction rate, (for example 11% in a 10 year follow up of UK healthy men [920]. The addition of genetic markers to the current panel of classical risk factors is an area of potential promise for clinical application [907]. Although, individually, the impact of any one genotype on risk is modest, when such risk-genotypes are common their combination may have a strong predictive power. Modeling has suggested [921] that only around 20 genes are usually needed to explain 50% of the burden of a disease in the population if the predisposing genotypes are common (> 25%), even if the individual risk ratios are relatively small (i.e. increasing risk by only 20-50%). Thus the combination of 10-20 meta-analyses confirmed SNPs may have good clinical utility, as has been suggested recently [922]. Recently a combination of SNPs in three meta-analyses proven genes has been reported to be of utility in identifying individuals at risk of type 2 diabetes, [923] confirming the potential of such an approach.

15.5 Pharmocogenetics

Much has been written about the potential of pharmacogenetic testing to inform therapy based on an individual's genetic makeup, and to decide the most effective choice of available drugs, or to avoid dangerous side effects. Currently, there is little hard data for either in the field of cardiovascular disease. The usual approach has been opportunistic use of drug trials in unrelated patients, and to look for differences in response or outcome by ‘candidate gene’ genotype, for example genes coding for drug metabolizing enzymes (activators and metabolizers), and enzymes and receptors involved in lipid metabolism, adrenergic response, etc. As with all association studies, initially promising results have often failed the test of replication in larger studies, and the relationship between the CETP Taq-I variant and response to statins has now been disproved [925]. The strongest data to date is the report [926] of a poorer cholesterol-lowering response to Pravastatin in the 7% of patients carrying a certain haplotype of the HMG CoA reductase gene (14% fall versus 19%), but if this is overcome simply by a higher dose, it is of little clinical relevance. Currently, the best example of avoiding side effects is determining genotype at the CYP2C9 locus with respect to warfarin treatment, since carriers for functional variants (> 20% of the population) require lower doses for optimal anticoagulation, and homozygotes, although rare, may well experience serious bleeding if given a usual dose [927]. The full potential of this field will only be realised with much further work.

15.6 Severe familial dyslipidaemias and CHD

There are many extremely rare inherited conditions where plasma lipids are abnormal and CHD risk is altered. Here we focus on only the three most common of these.

15.7 Familial hypercholesterolaemia (FH)

FH has an estimated prevalence of 1/500, [928] but may be much higher in some populations which recently increased in size (e.g. French Canadians, Afrikaners and Lebanese), as a consequence of the so-called founder effect. It is characterized by hypercholesterolaemia due to elevated plasma LDL levels, xanthomas, premature CHD and a family history of one or more of these. Angina, heart attacks or death typically occur in men between 30 and 50 years, and in women between 50 and 70 years, [929] and those who smoke, are hypertensive or have other risk factors are at particularly high risk. Several slightly different methods or scoring systems have been developed for the clinical diagnosis of definite, possible or suspected FH patients, [930–932] each have strengths and weaknesses, and have different sensitivity and specificity compared to the ‘gold standard’ of the presence of a detectable disease-causing DNA change [933]. It is likely that a combination of clinical, biochemical and DNA methods will give the highest clinical utility.

FH is present in 5-10% of individuals who develop CHD under the age of 55 years [929]. Thus the early identification of FH individuals will allow changes in lifestyle, including dietary intervention and smoking cessation advice as well as for drug treatment, and these measures, particularly statin treatment, will lead to a longer healthier life [934]. Statin therapy is warranted even in young FH individuals who currently have no evidence of CHD because of their high lifetime risk. Cost-benefit modeling has demonstrated the effectiveness of cascade testing in the relatives of FH patients, [935, 936] and an active program in The Netherlands has been particularly successful in identifying FH relatives in this way [930]. Similar programs are underway in several countries in Europe [937–940].

FH is an autosomal dominant inherited disorder and is usually caused by a mutation in the low-density lipoprotein receptor gene (LDLR). To date over 700 different mutations have been identified world-wide (see http://www.ucl.ac.uk/fh) although the spectrum within a single country is much smaller [940, 941]. Screening for deletions and rearrangements of the LDLR gene using a technique called multiplex ligation-dependent probe amplification (MLPA) [942] has become available, and it is known that up to 5% of FH patients may have such a deletion [943].

It is now known that mutations in at least two other genes can also cause similar symptoms and in approximately 3% of FH patients in the United Kingdom, North Europe and the USA a defect has been detected in the apolipoprotein B-100 gene (apoB), the ligand for the LDL-receptor. This disorder has been designated familial defective apolipoprotein B-100 (FDB) [944]. FDB appears to be somewhat milder in its expression that LDLR but hypercholesterolaemia occurs in childhood, and early CHD is frequent. Recently, defects in a third gene causing monogenic hypercholesterolaemia have been identified [945]. The gene called protein conver-tase subtilisin/kexin type 9 (PCSK9), codes for an enzyme that has also been called ‘neural apoptosis regulated convertase 1’ (NARC-1) which has recently been proposed to be involved in degrading the LDL-receptor protein in the lysosome of the cell and preventing it recycling [946]. Gain of function mutations in the PCSK9 gene could therefore cause increased degradation of LDL-receptors, reduced numbers of receptors on the surface of the cell and monogenic hypercholesterolaemia. An alternative mechanism for the hypercholesterolaemic effect has also been proposed, whereby the gain of function causes increased secretion of apoB-containing lipoproteins from the liver, with this being supported by in vivo turnover studies in patients carrying PCSK9 missense mutations [947] and by in vitro studies in transiently transfected rat liver cells [948]. One mutation in this gene, p.D374Y, has been reported in several independent families [945, 948–951] and appears to be associated with particularly high levels of untreated LDL-cholesterol levels and severe and early CHD, and poor response to statins. Finally, a recessive form of hypercholesterolaemia has been reported, caused by defects in a chaperone protein [952]. The frequency of this is unknown but it appears to be rare.

Since cholesterol levels in FH and non-FH subjects overlap, molecular genetic testing can be useful in the correct diagnosis of relatives in such families [953, 954].

Using currently available routine clinical genetic diagnostic techniques [931, 941], it is possible to demonstrate a mutation in the LDLR, PCSK9 or APOB gene in many of these patients, but this is usually only available in a research setting. Such specialist services are available in several European countries [937–941], but each country should have its own programme for genetic testing for FH because the spectrum of mutations varies between countries. Current data strongly suggests that DNA testing for FH complements cholesterol measurement in cascade screening to identify affected individuals unambiguously [955, 956]. However, an LDLR mutation is not found in all patients with a clinical diagnosis of FH, with reported detection rates being in the range of 60-90% depending to some extent on the molecular techniques employed, and the patients selection criteria. Detection rate is highest in those with a diagnosis of definite FH and lowest in those with only suspected FH [933, 941].

Although patients with no identified mutation may have a monogenic cause of the disorder in a yet-to-be discovered gene, it is also possible that some may have polygenic hypercholesterolaemia and have been misclassified using current clinical diagnostic criteria. These individuals would be expected to have a milder degree of hyper-lipidaemia, possibly not present from birth but only developing in later life, and would therefore be predicted to have a lower risk of CHD. In support of this, the CHD mortality risk in possible FH patients has been reported to be lower than in definite FH [957].

Current data supports the implementation of cascade testing for FH as being feasible and cost-effective but national implementation is limited to a small number of countries [956]. Funding and the infrastructure to support it may be the major stumbling blocks in implementing this in many countries. Concerns about the ethics of carrying out cascade testing, [958] and the potential psychological damage of cascade testing appear to have been largely dealt with [956].

15.8 Familial combined hyperlipidaemia (FCH)

This is the most common of the severe hyperlipidaemias, with a prevalence of perhaps 1/100 [959]. The genetic inheritance pattern is not as clear-cut as seen for FH, and a FCH is likely to be more polygenic/multifactorial than FH, but the identification of the gene(s) involved in a particular patient is likely to be of clinical utility, whether the disorder is caused by a ‘major gene’ or the interplay of several. Recently a major gene determining the FCH phenotype has been found in Finnish families, identified as the gene for Upstream Regulatory Factor 1 (USF1) – a major controller of lipid and glucose homeostasis [960]. In the liver, USF1 regulates the glucose-responsive expression of fatty acid synthase (FAS), a key enzyme in lipogenesis. USF1 also functions in adipose tissue and the pancreas, two organs heavily involved in metabolic homeostasis. Despite no specific mutation within USF1 being identified in FCH patients, a common haplotype composed of several SNPs is associated with risk of developing FCH [961]. To what extent these findings can be extrapolated to other populations is currently unclear [962]. Modifying genes, especially influencing the high triglyceride trait, include APOC3 and APOA5, [962] the latter representing a downstream target of USF1 and implying a USF1-dependent pathway in the molecular pathogenesis of dyslipidaemias.

15.9 Familial HDL deficiency Syndromes

Since the inverse and independent association between HDL-cholesterol and the risk of fatal and nonfatal CHD events has been established by clinical and epidemiological studies, low HDL cholesterol, most frequently defined as an HDL-cholesterol level less than 35 mg/dl (∼ 0.9 mmol/l), has become part of the multiparametric algorithms used for cardiovascular risk estimation [896, 919]. Familial defects that cause particularly low HDL levels, although rare, are therefore clinically important. Patients with HDL-cholesterol levels below the 5th percentile within a given population can be assumed to have Monogenic High Density Lipoprotein (HDL) deficiency [963]. These include defects in the genes of apolipoprotein A-I (apoA-I), adenosine tripho-sphate binding cassette transporter A1 (ABCA1) or lecithin:cholesterol acyltransferase (LCAT). Definitive diagnosis requires specialized biochemical tests and the demonstration of a functionally-relevant mutation in the appropriate candidate genes [963].

15.10 Coagulation disorders

Although familial monogenic disorders have been identified they are mostly extremely rare. However, more common mutations in the genes for clotting factor V (R506Q designated factor V Leiden) and for prothrombin (G20210A) have been identified, each with a carrier frequency of 2-3%, In a recent meta-analysis the perallele relative risks (RR) for coronary disease of factor V 1691A and of prothrombin 20210A were 1.17 (95% CI 1.08-1.28) and 1.31 (1.12-1.52), respectively [895].

16.1 Early detection of CVD in asymptomatic patients with MRI

MRI has been evaluated as a means of assessing the presence or absence of coronary artery stenosis. The value of this technique in detecting coronary artery stenosis is still in question [964–968]. Although the coronary arteries can be visualized in certain patients with the use of MRI, the sensitivity, specificity and robustness of this technique is not high enough to perform screening for coronary stenoses in asymptomatic people.

A potentially more useful approach for risk stratification is to perform in vivo imaging of the arterial wall using MRI. In vitro, MRI is able to differentiate between the plaque components of carotid, aortic, and coronary artery specimens obtained at autopsy [969–971]. Moreover, it has become possible to noninvasively depict coronary plaques by MRI [972–976]. Using optimized 3D imaging sequences to improve contrast between lumen and vessel wall, a spatial resolution of 0.66 × 0.66 × 2 mm³ can be obtained [974]. Regression of the lipid component of atherosclerotic plaques induced in animal models can now be demonstrated by serial in vivo MR examinations [977]. The current fast technical improvement has led to three-dimensional black blood vessel wall imaging which permits in vivo distinction between ‘normal’ and diseased vessel walls [978, 979]. Carotid, aortic and even coronary plaque assessment with MRI may soon lead to its use as a screening tool for quantifying subclinical disease, predicting future cardiovascular events and evaluating therapeutic interventions. For the present moment MRI is a promising research tool, but its use is limited to only a small number of research laboratories at this time. Thus, MRI is not yet appropriate for use in identifying patients at high risk for CAD. The Prevention Conference V participants [980] have recommended that more studies of MRI in CHD risk prediction should be encouraged. Additional technical development in this area is expected and should be of considerable value in the application of this emerging technology.

16.2 Quantitative assessment of coronary calcifications for the detection of asymptomatic high risk individuals

Coronary calcifications represent atherosclerosis of coronary arteries. Normally, they occur exclusively as atherosclerotic lesions within the intima layer and are not found in healthy coronary vessel walls [981–984]. On the other hand, atherosclerotic diseased coronary arteries do not necessarily always show calcifications. The extent of coronary calcifications correlates with the extent of the total coronary plaque burden [981, 982, 985]. It should be noted that coronary calcification is neither an indicator for stability nor instability of an atherosclerotic plaque [986–992]. The inflammatory component has been emphasized for patients with acute coronary syndrome, [993] underlining the concept of evaluation of the total coronary plaque burden by quantification of coronary calcium burden [988, 994].

The visualization of coronary calcium by means of fluoroscopy is not sensitive enough to detect clinically relevant information at early stages [995]. Recent developments in technology of the classic CT resulted in multislice CT-devices [996]. With the use of MS-CT it is possible to obtain a clear definition of the coronary vessels in most patients. However, the highest value of this technique seems to be its negative predictive value, reaching close to 98% in some studies. This very high negative predictive value of the technique leads to the consideration of using MS-CT for screening in certain subsets of high risk population. Still we need prospective studies to clearly determine which population may benefit most from this technology. In addition to that, coronary calcium score may be also reconsidered since the current technology provides more detailed information than was available in past years. If coronary calcium scanning is applied inappropriately, the proof of coronary calcium may lead to an unnecessary increase of diagnostic cardiac catheterizations or even coronary interventions in asymptomatic persons. Therefore, even in the presence of coronary calcium with its prognostic implications, the decision for coronary angiography remains unchanged and depends on the presence of angina pectoris and/or objective myocardial ischaemia [997–1012].

The Agatston score is the most widely used method of quantifying calcium on EBCT scan. Although today the prognostic impact of the Agatston score has been proven and accepted, one has to ask the key question whether the Agatston score is independent of the classical risk factors, meaning that the Agatston score provides additional clinically relevant information [994, 1013–1016]. Some publications supply an answer to this question: the Agatston score is an independent risk factor regarding the extent of coronary artery disease [1013, 1017–1022] and regarding prognostic impact; [997, 1001, 1021, 1023] for example, two men of the same age and identical classical risk profile may show an eightfold different risk of a coronary event [1023]. Aside from one study in older, high-risk patients, [1000] all studies showed that the prognostic value of coronary calcium offers information beyond the conventional risk factors [990, 1001, 1002, 1024]. The Rotterdam calcification study showed that the upper percentile range reflects a 12-fold increased risk of myocardial infarction – independent of the classical risk factors – even in elderly people [1025]. The calcium score also adds important prognostic information to the measurement of C-reactive protein [1026]. Furthermore, the extent of coronary calcium seems to reflect genetic components [1027]. Even psychosocial factors seem to play a role in the presence and extent of coronary calcification [1028, 1029].

16.3 Carotid ultrasound

Population based studies have shown a correlation between the severity of atherosclerosis in one arterial territory and involvement of other arteries [526]. The detection of atherosclerotic lesions in legs or carotid arteries is more accessible for noninvasive examinations than coronary or intra-cerebral arteries. Therefore, early detection of arterial disease in apparently healthy individuals has also focused on the peripheral arterial territory and on the carotid arteries.

Sonography of superficial arteries is a relatively inexpensive means of noninvasively visualizing the lumen and walls of arteries which are involved in the ubiquitous process of atherosclerosis. Risk assessment using carotid ultrasound focuses on measurement of the intima-media thickness (IMT) and plaque characteristics.

16.4 Ankle-brachial index (ABI)

16.5 Ophthalmoscopy for atherosclerosis screening

Recently it has been shown that the extent of retinal arteries atherosclerosis correlates with the extent of the total coronary plaque burden. Atherosclerosis of retinal arteries also strongly correlates with plasma total cholesterol, LDL-cholesterol, triglycerides and apoprotein B levels. Since ophthalmoscopy is a noninvasive technique, easy to perform and has no adverse effects it might be used to detect asymptomatic nonhypertonic individuals at high risk for cardiovascular events [1047, 1048].

19.1 Antiplatelet therapies

19.2 Beta-blockers

In a meta-analysis of β-blockers following myocardial infarction, there was evidence of a significant reduction in all-cause mortality, cardiovascular death and in particular sudden cardiac death, as well as nonfatal reinfarction [629]. The benefits of β-blockade are greatest in older patients (more than 60 years) and in patients at increased risk of reinfarction and death (e.g. patients with left ventricular (LV) dysfunction or arrhythmias or both). Apart from the indication in postinfarction, β-blockers have also comprehensive evidence from multiple large-scale trials to reduce all-cause mortality in patients with heart failure due to any cause, including coronary heart disease (US-Carvedilol; COMET; CIBIS, MERIT-HF trials) [1085–1088].

Taken together, β-blockers are indicated, providing there are no contraindications, (i) in the treatment of heart failure, (ii) as prophylaxis following myocardial infarction, including patients with diabetes; (iii) to relieve symptoms of myocardial ischaemia; and (iv) to lower blood pressure to the goal of less than 140/90 mmHg, except in diabetic patients where alternative classes of antihypertensive drugs can be considered before β-blockers.

19.3 ACE inhibitors

19.4 Angiotensin-receptor blockers (ARBs)

Generally speaking ARBs are indicated in all patients who have an indication for ACE inhibitor therapy, but cannot tolerate ACE-inibitors, for example, due to side effects. In addition ARBs in combination with ACE inhibitors can reduce morbidity (i.e. rate of rehospitalization) in patients suffering from congestive heart failure (VAL-HEFT, CHARM trials) [1094, 1095].

19.5 Calcium channel blockers (CCBs)

This drug class has been shown to reduce cardiovascular outcomes in people with arterial hypertension.

In post-MI patients with contraindications to β-blockers and no evidence of heart failure, verapamil may be considered based on the results of a single large clinical trial (DAVIT-trials) [1096].

Hence, calcium channel blockers are indicated for the reason of reducing blood pressure to target less than 140/90 mmHg or less than 130/80 mmHg (diabetes).

OPEN IN VIEWER

Table 18

Indications for antithrombotic therapy in patients with atrial fibrillation

	Risk category	Recommended therapy
No risk factors	Aspirin, 81 to 325 mg daily
One moderate-risk factor	Aspirin, 81 to 325 mg daily, or warfarin (INR 2.0 to 3.0, target 2.5)
Any high-risk factor or more than 1 moderate-risk factor	Warfarin (INR 2.0 to 3.0, target 2.5)a
Less validated or weaker risk factors	Moderate-risk factors	High-risk factors
Female gender	Age greater than or equal to 75 y	Previous stroke, TIA or embolism
Age 65 to 74 y	Hypertension	Mitral stenosis
Coronary artery disease	Heart failure	Prosthetic heart valvea
Thyrotoxicosis	LV ejection fraction 35% or less
	Diabetes mellitus

INR, international normalized ratio; LV, left ventricular; TIA, transient ischaemic attack. ^aIf mechanical valve, target international normalized ratio (INR) greater than 2.5.

19.6 Diuretics

Diuretics are indicated for the following reason: to reduce blood pressure to target less than 140/90 mmHg. Thiazide diuretics are not recommended as first-line antihypertensive agents in diabetic patients or those at high risk of developing type 2 diabetes.

19.7 Anticoagulation

Systemic anticoagulation with coumarins is generally not indicated prophylactically in patients with coronary artery disease. However, anticoagulation can be considered in selected patients following myocardial infarction at increased risk of thrombo-embolic events including patients with large anterior myocardial infarction, left ventricular aneurysm or thrombus, paroxysmal tachyarrhythmias and chronic heart failure, particularly in combination with aspirin (WARIS-II trial) [1097]. In patients with paroxysmal or permanent atrial fibrillation, systemic anticoagulation is indicated as shown in Table 18 [1098].

In patients with coronary heart disease, or other cardiac disease, systemic anticoagulation is indicated for the following reasons:

20.1 Defining the problem

Several studies have been made to evaluate the effect of different implementation strategies of guidelines on clinical practice.

EUROASPIRE I (1995/96) [1099] and II (2000/01) [101] surveys, both showed a high prevalence of unhealthy lifestyles, modifiable risk factors and inadequate use of drug therapies to achieve blood pressure and lipid goals in patients with established CHD, with wide variations in medical practice between countries. The comparison of EUROASPIRE I and II surveys showed adverse lifestyle trends, especially the substantial increase in obesity in every country, and smoking among younger patients. Half of patients continued smoking after a coronary event. There was no change between surveys in the proportion of patients who reached their blood pressure goal of less than 140/90 mmHg and one in two patients with established CHD are still not reaching this goal, despite the real improvement in the proportion of patients who reach the total cholesterol target (Tables 19 and 20).

Although implementation of guidelines improved over time, subgroups are still not receiving appropriate therapy, especially patients with stable angina and the elderly who are under treated. Many surveys have shown similar results; some also showing great variance between countries [1100–1102].

There is considerable potential through Europe in coronary patients and their families to raise the standard of preventive cardiology through more lifestyle intervention, control of other risk factors and optimal use of prophylactic drug therapies in order to reduce the risk of recurrent disease and death. Similar results have been shown for stroke.

OPEN IN VIEWER

Table 19

Main findings in EUROASPIRE I and II: risk factors

	EASP I (%)	EASP II (%)
BMI ≥ 25	78	81
BMI ≥ 30	25	33
BP ≥ 140/90	55	50
Cholesterol ≥ 5	67	59
Cigarettes	19	21

BMI, body mass index; BP, blood pressure.

OPEN IN VIEWER

Table 20

Main findings in EUROASPIRE I and II: drug use

	EASP I (%)	EASP II (%)
ASA	81	84
Beta blocker	54	66
ACE inhibitor	30	43
Ca Antagonist	36	26
Lipid lowering	32	63
Statin	19	58

ASA, acetylsalicylic acid.

Traditional approaches to improve uptake of research findings have focused on better availability and presentation of evidence by identifying, synthesizing, and disseminating evidence to doctors in practical accessible formats – e.g. reviews in clinical journals, clinical guidelines, better access to electronic sources of information, continuing medical education (CME) courses, and conferences. Although this strategy may be all that is needed to ensure the uptake of some simple changes, most innovations require further efforts. Most clinicians can barely keep pace with the rapid advances in healthcare knowledge. Shaneyfelt calculated that general internists would need to read 20 articles a day all year round to maintain present knowledge [1103]. Although the availability of systematic reviews and guidelines reduces the need for doctors to read original studies, they still find it difficult to keep up with such syntheses. Even if doctors are aware of the evidence and are willing to change, to alter well established patterns of care is difficult, especially if the clinical environment is not conducive to change.

20.2 Barriers to the implementation of guidelines

It is essential that clinical guidelines are in concordance with priorities in the health system and with ethical values most clinicians can agree upon. If not, this may be an important reason why many clinicians do not follow guidelines [1104].

The implementation of these guidelines should be based on national surveys to adjust them to the stratification of risk factors and premature CVD death in the individual country and bring them in accordance with priorities set by the health authorities and the professional bodies. The workload put on the health system should be affordable and should not imply that resources should be allocated to prevention strategies when the outcome for the population is better by alternative use.

Given agreement that the implementation of prevention is a priority, the next step in the implementation of these guidelines is the involvement of the clinicians, in primary and secondary care.

Analyses of the barriers to changing practice, such as a review of 76 doctors, have shown that obstacles to change in practice can arise at different stages in the healthcare organization, or the wider environment. Most theories on implementation of evidence in healthcare emphasize the importance of developing a good understanding of such obstacles to develop an effective intervention (Table 21).

20.3 Doctor-patient relationship

The preventive interventions must be based on a patient-centered approach, where the doctor pays full attention to appraise and meet the patient's concerns, beliefs and values, and respects the patient's choice even if it is not in concordance with the doctor's first proposal. The changing of lifestyle or taking medication often means for the rest of the patient's life, so the decision must be owned by the patient. Therefore, treatment goals should be set in collaboration with the patient, taking into account the values and priorities of the patient. If the treatment goals are unaffordable, it may lead to frustration and clinical neglect, both by the doctor and the patient. The doctor should explore the patient's important values, beliefs and expectations regarding the prevention measures to be taken.

20.3.1 Physician-related methods to improve implementation

It has been argued that the application of guidelines in a setting of rigorous control gives the best chances to improve clinical practice. However, most front-line clinicians work in settings where such control is not practical, and mostly not wanted. The very character of this specific task – primary and secondary prevention – is not suitable for this strategy.

OPEN IN VIEWER

Table 21

Barriers to implementation of guidelines

Practice related (organizational context)

•Financial disincentives (e.g. lack of reimbursement)

•Organizational constraints (e.g. lack of time)

•Doctor-patient relationship (e.g. patient's expectations, doctor's perception of…)

Prevailing opinion (social context)

•Opinion leaders (e.g. persons not agreeing with the guidelines)

•Lack of medical training

•Advocacy (e.g. by pharmaceutical companies)

Knowledge and attitudes (professional context)

•Sense of competence (e.g. competence in risk appraisal)

•Clinical uncertainty (e.g. overtreatment, unnecessary investigations)

•Information overload (e.g. inability to appraise evidence)

Audit and feedback, where practitioners are given data on their performance is another tool that has been used to improve practice. It seems logical that practitioners will change their practice when they get feedback indicating that their present practice is inconsistent with that of their peers or with evidence based guidelines. A recent meta-analysis from the Cochrane collaboration shows that the effect of audit and feedback, if any, is modest [1105].

The decision to start preventive measures and follow them up is more value-related than treatment of established disease, so the values and attitudes of the doctors and the patients are more important. In addition, most clinical decisions are taken more intuitively, on the basis of recognition patterns or other internal mental shortcuts (heuristics) of the individual doctor. How this affects the application of guidelines is not known, and more research is needed.

20.4 Important arenas for training

There is a need for training of doctors in patient-centered preventive care, with emphasis on:

20.5 Implementation strategies

The implementation strategies should consist of a package of different measures, working in combination: