The search for new risk factors for coronary heart disease: occupational therapy for epidemiologists?

Robert Beagleholea and Paul Magnusb

a University of Auckland, Auckland, New Zealand (on leave).
b Australian Institute of Health and Welfare, GPO Box 570, Canberra, ACT 2601, Australia.

Abstract

The identification of the proximal causes of coronary heart disease (CHD) during the second half of the 20th century contributed to the prevention of premature CHD and the extension of life expectancy in middle-aged and older people in many wealthy countries. These major CHD risk factors—high blood cholesterol, high blood pressure, cigarette smoking and physical inactivity—satisfy public health criteria of causality. Strong epidemiological evidence suggests that they explain at least 75% of new cases of CHD. However, the search for ‘new’ or ‘emerging’ CHD risk factors continues, partly justified by a myth that minimizes the contribution of the major risk factors.

The public health criteria of causality were applied to the following proposed new risk factors: thrombotic factors and serum homocysteine levels; infectious agents; early life exposures including prenatal factors; genetic influences; oestrogen deficiency; and the role of the psychosocial environment. None of these factors are as important as the established risk factors for epidemic CHD and their potential contribution for improving population health is limited or unclear. Research into unexplained variations in the occurrence of CHD and into life course influences and socioeconomic inequalities may provide extra leads to effective public health action. Especially important is research on the upstream social and economic determinants of CHD and its major risk factors, on the spread of the CHD epidemic to poorer populations, and into prevention policy and programme effectiveness. Available evidence supports the feasibility and effectiveness of population-wide prevention directed towards increasing the proportion of people at low risk of CHD. The vast majority of the public health effort should be directed to this approach rather than to the high risk individual approach. There is still a major gap between knowledge and action in preventing the CHD epidemics.

Accepted 15 March 2002

The main approach to reducing the social and economic burden due to coronary heart disease (CHD) and other non-communicable diseases is based on the identification of risk factors at the level of the individual.1 This work began with the Framingham study in 19482 and has continued unabated. Epidemiological studies have identified the major causes of CHD and contributed to important public health gains, notably the prevention of much premature CHD and the extension of life expectancy in middle-aged and older people in many wealthy countries.3 One of the earliest studies, the Seven Countries Study, made important contributions to identifying population-level risks.4 Despite the power of current knowledge to explain the occurrence of CHD in populations, there is strong support for continuing the search for new CHD risk factors,1,5 although there is now debate about the limitations of the ‘risk factor’ approach.6

This paper has three parts. First, we summarize the explanatory power of existing knowledge on the causes of epidemic CHD at the country level and the potential for the primary prevention of epidemic CHD. Second, we assess the potential gain from research into ‘emerging’ or new CHD risk factors. Finally, we stress that the implementation of policies and programmes based on existing knowledge could significantly improve adult health in all but the poorest countries in a few years.

In assessing the potential contribution of new and established risk factors to the prevention of CHD, we use the following public health criteria. Do the risk factors offer the opportunity to intervene before symptoms or signs of end-organ damage occur? Do the associations with CHD satisfy evidence of causality with the totality of evidence being strong and consistent? Are the factors prevalent enough to be independently responsible for a significant part of the coronary epidemic? For ‘new?’ factors, there are two further questions that are rarely asked but which are key to assessing their public health importance: how much extra do they add to explaining the CHD epidemics (not just to variations within the epidemics); and do they suggest affordable strategies to improve CHD that are not already offered by the established factors?

What do we know about the causes of CHD?

This year CHD will be responsible for about 7 million of the 56 million total world-wide deaths.7 For demographic reasons, most of these deaths occur in the poorer regions of the world. It is expected that in 2020 CHD will be the leading cause of death and a leading cause of disability-adjusted life years lost. The burden of CHD varies by region; for example, it is much less common than stroke in East-Asian populations.

The underlying causes of the CHD epidemic are the society-wide economic and cultural factors that determine whether a diet with a high proportion of saturated fat and low in antioxidants becomes widespread.2,4 This diet leads to unfavourable lipid features and thus population-wide atherosclerosis, which is the basis of the CHD epidemics. The prevailing social and economic conditions are also responsible for the emergence and widespread distribution of other important contributing causes: tobacco smoking, physical inactivity and other inappropriate aspects of diet, with the latter two interacting to produce excess weight and high levels of blood pressure. These major causes have a close and precisely defined (proximal) relationship to the CHD epidemics and are well established scientifically. They have a strong, independent, dose-related, and biologically plausible association with CHD as shown by a full range of laboratory, clinical, and observational and experimental epidemiological studies conducted in many parts of the world. For over 25 years there has been evidence that these factors account for most of the coronary epidemics, at least in a typical Western society such as the US.8 For example, evidence from cohort studies and randomized trials indicates that after adjusting for regression dilution biases, a long term change of 0.6 mmol/l in serum cholesterol concentration among middle-aged men corresponds to a coronary risk change of at least 25%.9 Similarly, every 5 mmHg change in usual diastolic blood pressure corresponds to a 21% change in coronary risk.10 Evidence from cohort studies and observations of countries with very low CHD death rates suggest that optimal population levels for coronary health are <=4 mmol/l for cholesterol and <=70 mmHg for diastolic blood pressure. On average, cigarette smoking increases the risk of CHD death by 70% compared with not smoking.11 Physical inactivity may be under-rated as a critical risk factor for CHD, at least in part because of the absence of clinical trial data;12 however, there is no doubt that changing activity patterns are contributing to the rapid development of obesity as a major public health problem in most countries.13

In a given individual, the major risk factors do not act as simple additive risks; they interact synergistically in various combinations to increase total CHD risk in a complex way. From a public health perspective, the combined contribution of well established risk factors to the burden of CHD can be estimated by comparing event rates in a low risk group that has favourable (preferably optimal) levels of risk factors with the rates in the remainder of the same population. Several large cohort studies have examined the combined impact of the established risk factors or factors closely related to them. For example, from 1975 data summarizing the combined 10-year follow-up of eight major US prospective studies in the National Cooperative Pooling Project,8 it was calculated that 66% of CHD events among 30- to 69-year-old men were attributable to high serum cholesterol (defined as >=6.5 mmol/l), high blood pressure (diastolic level >=90 mmHg) and current use of cigarettes.14 A similar population attributable fraction was obtained in an unpublished analysis of the Whitehall I Study based on cigarette smoking and the lowest quintiles of blood pressure and cholesterol.14 The most authoritative analysis came in 1986 when Stamler et al. published their 6-year follow up of 356 222 men who were aged 35 to 57 when screened for entry into the US Multiple Risk Factor Intervention Trial. Using the same above optimal cut-point as the Pooling Project for blood pressure but a lower one of >=4.71 mmol/l for blood cholesterol, they found that 75% of CHD deaths could be attributed to the three classical risk factors.15 Longer follow-ups of the MRFIT cohort in 199216 and 1999,17 with the addition of other large male and female cohorts, strongly reaffirmed the earlier findings.

Complementary evidence comes from studies of Harvard University alumni and of US nurses. From a 9-year follow-up of the self-reports of 10 269 men who were aged 45 to 84 years in 1977, it was estimated that 58% of coronary deaths could be attributed to cigarette smoking, not engaging in moderately vigorous sports, physician diagnosed hypertension, overweight (body mass index [BMI] >= 26) or a history of early parental death.18 From a 14-year follow-up of the validated self-reports of 84 129 women in the Nurses Health Study who were aged 34 to 59 in 1980, 82% of coronary events were attributed to cigarette smoking, physical inactivity (average less than half an hour daily of moderate or vigorous activity), overweight (BMI >= 25), or being in the less favourable 60% of the cohort based on a composite diet score (derived from intake of cereal fibre, marine n-3 fatty acids and folate; and from PS ratio, level of trans fat intake and glycaemic load).19

Together these studies present a coherent picture of a very large contribution from the established risk factors to the CHD epidemic. This is especially impressive considering that none of the studies covered all the major risk factors or used uniformly optimal cut-points for them. Furthermore, since the causes of CHD are qualitatively the same in all populations, and even the quantitative associations are remarkably similar, for example, between populations in Asia and western countries,20 it is likely that estimates of the contribution of the established risk factors to the emerging epidemics in developing countries will be similar.

In summary, the major risk factors of inappropriate diet and physical inactivity (primarily as expressed through unfavourable lipid concentrations, high BMI, and raised blood pressure), together with tobacco use, explain at least 75% of new cases of CHD. In the absence of these risk factors, CHD is a rare cause of death. The optimal levels of CHD risk factors are known; unfortunately, only about 5% of adult men and women in wealthy countries are at low risk with optimum risk factor levels.17,19

The search for new risk factors

The justification for the search for new risk factors has for decades depended heavily on a claim that has turned out to be nothing more than a myth: the ‘50% myth’,14 namely that the established coronary risk factors explain at most half of the occurrence of CHD. The implication is that there is ‘another 50%’, or more, yet to be discovered, hence the supposed justification for more research into new risk factors. In fact, this claim is based on simple and repeated assertions backed up by inappropriate citations or secondary quotations, but not by plausible empirical findings. The 50% myth has been remarkably persistent and widely accepted despite the findings of major studies that should have seriously called it into question long ago.8,15,17 It has been most convenient for those wishing to justify their search for new risk factors.

This search is a preoccupation of epidemiologists and other researchers. It follows six main areas: thrombotic factors and the effect of biochemical markers, for example, serum homocysteine levels; the role of infectious agents; the influence of early life exposures including prenatal factors; the contribution of multiple genes; oestrogen deficiency; and the role of the psychosocial environment, variously defined.

Thrombotic factors are of great importance in determining the clinical expression of CHD and thus for determining pharmacological therapy. However, from a practical perspective, knowledge of their importance does not alter approaches to primary prevention based on the other major risk factors.21 The relationship of serum homocysteine levels to increased CHD risk has been extensively studied in populations in wealthy countries and trials of folate supplementation are assessing the effects of reducing homocysteine levels. In the prospective British Regional Heart Study, total homocysteine concentrations >16.6 µmol/l accounted for 13% of the attributable risk of myocardial infarction;22 however, the results of the prospective studies are not consistent and are at best modest.23 Furthermore, it appears that a sensible population-wide intervention to reduce homocysteine levels will be the promotion of a diet that is already indicated for many other reasons.24

Despite research over several decades and the examination of multiple markers, evidence of a causal relationship between inflammation and CHD is lacking. One of the most investigated markers is C-reactive protein although it remains uncertain whether this marker is an independent risk factor for CHD.25 A recent meta-analysis of the association of chronic chlamydia pneumonia infection and CHD reliably excludes any strong association.26 Other potential bacterial and viral causes of CHD that continue to be investigated with disappointing results include Helicobacter pylori27 as well as general inflammatory biomarkers. A recent meta-analysis of soluble adhesion molecules which help recruit circulating leucocytes to sites of inflammation and are general markers of inflammation, concluded that these markers do not add much predictive information to that provided by the established risk factors,28 though these findings may not diminish the enthusiasm for further research in this area.29

Genetic research might improve our understanding of individual susceptibility to disease and thus to risk management in high risk individuals,30 but it will not substantially contribute to cardiovascular disease prevention at the population level. The major population trends and differences in the occurrence of CHD are due to environmental factors and the different stages in the evolution of cardiovascular disease epidemics.31 The doubling of the prevalence of obesity in the UK over two decades is an example of the importance of the environmental determinants of chronic disease, even if genetic susceptibility is involved in determining which individuals will be most affected.13 A fundamental point is that genetic factors cannot explain the rapid increases and decreases in population CHD rates witnessed over the last three decades.

Genetic factors also provide the strongest example of a major adverse public health effect of the continuing search for new risk factors, namely the emphasis on categorizing individuals according to personal disease risk.32 Everything we know about risk factors indicates that they have a very poor ability to discriminate between individuals who will develop CHD and those who will not. Even with smoking, with its high relative risk of subsequent lung cancer, much higher than for CHD, at least one half of regular smokers will not die as a result of their habit.33

Research on the possible prenatal and early determinants of CHD and the programming hypothesis has led to important insights, for example, on the need to adopt a life course approach to the control of chronic diseases.34,35 However, the contradictory results of this research on the Barker hypothesis, the difficulties in eliminating confounding and establishing causal relationships between early life events and later CHD,36 and the obvious importance of later environmental influences on risk, are such that it is difficult to determine the unique public health significance of this hypothesis.

There is already ample justification for public health interventions to ensure good nutritional status among young women and for pregnant women this includes the direct health of the fetus and later of the infant. The rapid rise and fall of epidemic CHD and the reversal of the social class gradients of CHD in the third quarter of the last century in the UK37 argue strongly for the predominant role of the later environment in determining the modern CHD epidemics.

The history of the association between oestrogen deficiency/ hormone replacement therapy (HRT) and CHD is a reminder of the need for public health policy to be guided by solid evidence, especially when there is strong potential for confounding of the observed relationships. The early enthusiasm for HRT for CHD primary and secondary prevention is now giving way to much more cautious recommendations echoing the warnings from epidemiologists over a decade ago.38 The lack of concordance of the observational data with the limited results from clinical trials suggests the strong likelihood of a healthy person effect; that is, women taking HRT have lower risks of subsequent CHD for a variety of reasons.

Research on psychosocial factors in the aetiology of CHD has followed many different paths over the last five decades, from personality type to ‘job control’. Studies based on British civil servants suggest that job control may have an independent and dose-related association with a modest CHD gradient, both between and within employment grades.39 However, the concepts under discussion have been difficult to define and operationalize and the independent effect of these factors will presumably be weakened after the effects of the established risk factors are adjusted for regression dilution biases.10 Since most of this research has been conducted on selected populations and almost all in wealthy populations, it remains to be seen whether the findings apply to countries with lower CHD levels. Within populations some of these factors may have some explanatory power over and above the established risk factors. However, since causality is still unproven and practical interventions have not been developed, it would be premature to accord the psychosocial characteristics the status of risk factors or to estimate their overall contribution to population CHD levels.

The focus on individual risk status and the identification of high risk groups for targeted interventions continue to distract attention and resources from the more important public health goal of reducing the population risk of CHD. This point applies to the established risk factors as well, but in their case the risky behaviours are amenable to sustained and broad public health measures. In summary, none of the new or emerging factors yet meet the public health test of a risk factor for epidemic CHD and the related pay off for improving population health seems limited or unclear.

Discussion

We recognize the imperfections of the population attributable fraction approach; for example, the fractions for many chronic diseases can be increased by defining risk factors in such a way that nearly the entire population is labelled ‘at risk’.32 However, with CHD in western countries this is the case and the risk factor levels defined by optimal cut-points are realistic and achievable. Despite other limitations and the dichotomous definition of high blood pressure and high cholesterol, the population attributable fraction approach remains the only acceptable method of estimating the combined impact of the risk factors.14

We are not disparaging all research directed towards a greater understanding of the occurrence of CHD. To begin with, the search for more information does not have to be predicated on some kind of ‘percentage war’, although the 50% myth has been regrettably exploited in the past. Even if the major established risk factors appeared to ‘explain’ 100% of CHD, and despite the immense prospects they offer for prevention, they explain only the final common causal pathway for the epidemics. Research is required to explore the broader social and economic determinants of the major CHD risk factors.

We also recognize the limited ability of established risk factors to explain many variations in the occurrence of CHD, both within and especially among countries (although there are exceptions9,40). There are many interesting issues to explore in connection with these variations. The explanations are likely to be complex involving methodological issues, differing time courses of the epidemics,31 and complex interactions among the underlying causes.41 An important methodological issue is the impact of measurement error in underestimating the explanatory power of the established risk factors.10 Of particular interest is the life course approach of examining influences that accumulate over a lifetime,34 and complementary attempts to explain socioeconomic inequalities in risk, between both individuals and regions. Special attention is required to ensure that public health approaches to the established risk factors reduce socioeconomic inequalities in CHD, rather than exacerbate them.

New research is required on CHD prevention policy and programme effectiveness and on issues of importance for the spread of the CHD epidemics to poor populations. Policy directed research will have the biggest public health payoff in the short term, as it has had for tobacco control.33 The underlying causes of these epidemics lie in the social, economic and cultural domains, not easily investigated by traditional epidemiological methods. There is an urgent need for epidemiologists and other public health scientists to explore the applicability of new research methods to the upstream determinants of CHD, as is happening now with investigations into the health effects of global climate change.42 Some of the required research is more a matter of academic interest; some may in time provide extra leads to effective public health action. However, the unresolved issues should not detract from the urgency of applying what we know, especially in developed countries for which the evidence is incontrovertible.

Acting on what we know

While research on possible new risk factors and unresolved issues is at best only moderately promising, there is a major gap between what we know about the causes of CHD and our willingness to act on this knowledge and prevent a most preventable major epidemic.43 A striking deficiency has been in the development and application of appropriate prevention and control policies based on public health research evidence. The full application of existing prevention and control knowledge requires a major shift in the balance of preventive efforts to the population approach to primary prevention, especially in the poorer regions of the world. The global public health context is, of course, complex and presents difficult challenges for public health practitioners more used to dealing with risk factors at the level of the individual. However, this complexity must not become an excuse for inaction, especially since we have substantial evidence of the effectiveness of public health interventions, for example, the effect of co-ordinated and comprehensive tobacco control policies.

The critical policy question now, especially for poor countries, concerns the appropriate balance between primary and secondary prevention and between the population and high risk approach to primary prevention. If the goal is to significantly increase the proportion of the population at low risk status, the only strategy with this potential is the population-wide approach to primary prevention.44 All other strategies will, at best, only restrain the epidemics; they will not prevent them. The challenge is to implement this approach to primary prevention, that is, to shift the population risk factor distribution to the left. The ultimate public health goal is the reduction of population risk, and since 95% of the population, at least in developed countries, is not at the optimal risk level, it follows that the majority of prevention and control resources should be directed towards this goal. The remaining very small proportion of resources should go to the high risk individual approach.

Evidence is available in support of the policies required for the task of shifting risk distributions. The Asia Pacific Cohort Studies Collaboration indicates that a 2% reduction of mean blood pressure, achieved by a shift of the blood pressure distribution to the left, has by the year 2020 the potential to prevent annually 1.2 million stroke deaths (approximately 15% of all stroke deaths) and 0.6 million coronary deaths (6% of all CHD deaths).20 Reductions in mean blood pressure of this magnitude have been achieved in Australia,45 and could readily be achieved in many other populations, for example by reducing the salt content of manufactured food.46 Where manufactured food with its high salt content is not readily available, a major educational effort is required to reduce the amount of salt used in food preservation and cooking. Favourable shifts in the population distributions of abnormal blood lipid levels could be achieved by the wider adoption of healthy dietary patterns based on the traditional Mediterranean diet.47 Much more attention should also be directed to modifying the environmental determinants of physical inactivity and the resulting obesity.48 As a priority, it is important to build on the demonstrated success of comprehensive tobacco control policies.49

An important question is why the public health community has been reluctant to act fully on the available evidence on the causes of CHD. The answer probably lies with the dominant health paradigm that supports an individualistic approach to health improvement and ignores the wider social and economic determinants of the health of populations. It is time for public health practitioners to recommit to the fundamental justification for our profession—protecting and promoting the health of the populations. While not forgetting the great public health potential of acting on the major established risk factors, increased attention is also required to improve understanding of the social and economic determinants of the major risk factors and overall population health status.50

Acknowledgments

The authors acknowledge the helpful comments of Annette Dobson, Stan Bennett, Peter Whincup and Robert Clarke.

References

1 Nieto FJ. Cardiovascular disease and risk factor epidemiology: a look back at the epidemic of the 20th century. Am J Public Health 1999;89:292–94.[ISI][Medline]

2 Kannel WB. The Framingham Study: 50-year legacy and future promise. J Atheroscler Thromb 2000;6:60–66.[Medline]

3 Beaglehole R. International trends in coronary heart disease mortality and incidence rates. J Cardiovasc Risk 1999;6:63–68.[ISI][Medline]

4 Keys AB. Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press, 1980.

5 Lefkowitz RJ, Willerson JT. Prospects for cardiovascular research. JAMA 2001;285:581–87.[Abstract/Free Full Text]

6 McKinlay JB, Marceau LD. A tale of 3 tails. Am J Public Health 1999;89:295–98.[ISI][Medline]

7 World Health Organization. The World Health Report 2000. Health Systems: Improving Performance. Geneva: WHO, 2000.

8 Marmot M, Winkelstein W. Epidemiological observations on intervention trials for prevention of coronary heart disease. Am J Epidemiol 1975;101:177–81.[ISI][Medline]

9 Law MR, Wald NJ, Thompson SG. By how much and how quickly does reduction in serum cholesterol concentration lower risk of ischaemic heart disease? BMJ 1994;308:367–73.[Abstract/Free Full Text]

10 MacMahon S, Peto R, Cutler J et al. Blood pressure, stroke and coronary heart disease. Part 1. Prolonged differences in blood pressure: prospective observational studies corrected for the regression dilution bias. Lancet 1990;335:765–74.[ISI][Medline]

11 Department of Health and Human Services. Reducing the Health Consequences of Smoking: 25 Years of Progress: A Report of the Surgeon General. Washington, DC: Government Printing Office, 1989. (DHHS publication no. (CDC) 89-8411).

12 Paffenbarger RS, Blair S, Lee I-M. A history of physical activity and cardiovascular health and longevity: the scientific contributions of Jeremy Morris, DSc, DPH, FRCP. Int J Epidemiol 2001;30:1184–92.[Abstract/Free Full Text]

13 Prentice AM, Jebb SA. Obesity in Britain: gluttony or sloth? BMJ 1995;311:437–39.[Free Full Text]

14 Magnus P, Beaglehole R. The real contribution of the major risk factors to the coronary epidemics: time to end the ‘only 50%’ claim. Arch Int Med 2001;161:2657–60.[CrossRef][ISI]

15 Stamler J, Wentworth D, Neaton JD. Is relationship between serum cholesterol and risk of premature death from coronary heart disease continuous and graded? Findings in 356,222 primary screenees of the Multiple Risk Factor Intervention Trial (MRFIT). JAMA 1986; 256:2823–28.[Abstract]

16 Stamler J. Established major coronary risk factors. In: Marmot M, Elliott P (eds). Coronary Heart Disease Epidemiology. Oxford: Oxford University Press, 1992, pp.35–66.

17 Stamler J, Stamler R, Neaton JD et al. Low risk-factor profile and long-term cardiovascular and non-cardiovascular mortality and life expectancy. Findings for 5 large cohorts of young adult and middle-aged men and women. JAMA 1999;282:2012–18.[Abstract/Free Full Text]

18 Paffenbarger RS, Hyde RT, Wing AL, Lee I-M, Jung DL, Kampert JB. The association of changes in physical-activity level and other lifestyle characteristics with mortality among men. N Engl J Med 1993;328:538–45.[Abstract/Free Full Text]

19 Stampfer MJ, Hu FB, Manson JE et al. Primary prevention of coronary heart disease in women through diet and lifestyle. N Engl J Med 2000;343:16–22.[Abstract/Free Full Text]

20 Eastern Stroke and Coronary Heart Disease Collaborative Research Group Blood pressure, cholesterol, and stroke in eastern Asia. Lancet 1998;352:1801–07.[CrossRef][ISI][Medline]

21 Meade TW. Cardiovascular disease—linking pathology and epidemiology. Int J Epidemiol 2001;30:1179–83.[Abstract/Free Full Text]

22 Whincup PH, Refsum H, Perry IJ et al. Serum total homocysteine and coronary heart disease: prospective study in middle aged men. Heart 1999;82:448–54.[Abstract/Free Full Text]

23 Danesh J, Lewington S. Plasma homocysteine and coronary heart disease: systematic review of published epidemiological studies. J Cardiovasc Risk 1998;5:229–32.[Medline]

24 Riddell LJ, Chisholm A, Williams S et al. Dietary strategies for lowering homocysteine concentrations. Am J Clin Nutr 2000;71:1448–54.[Abstract/Free Full Text]

25 Danesh J, Whincup P, Walker M et al. Low grade inflammation and coronary heart disease: prospective study and updated meta-analysis. BMJ 2000;321:199–204.[Abstract/Free Full Text]

26 Danesh J, Whincup P, Walker M et al. Chlamydia pneumoniae IgG titres and coronary heart disease: prospective study and meta-analysis. BMJ 2000;321:208–13.[Abstract/Free Full Text]

27 Whincup P, Danesh J, Walker M et al. Prospective study of potentially virulent strains of Helicobacter pylori and coronary heart disease in middle-aged men. Circulation 2000;101:1647–52.[Abstract/Free Full Text]

28 Malik I, Danesh J, Whincup P et al. Soluble adhesion molecules and prediction of coronary heart disease: a prospective study and meta-analysis. Lancet 2001;358:971–75.[CrossRef][ISI][Medline]

29 Ridker PM. Role of inflammatory biomarkers in prediction of coronary heart disease. Lancet 2001;358:946–48.[CrossRef][ISI][Medline]

30 Day INM, Wilson DI. Genetics and cardiovascular risk. BMJ 2001;323: 1409–12.[Free Full Text]

31 Law M, Wald N. Why heart disease mortality is low in France: the time lag explanation. BMJ 1999;318:1471–80.[Free Full Text]

32 Rockhill B, Kawachi I, Colditz GA. Individual risk prediction and population-wide disease prevention. Epidemiol Rev 2000;22:176–80.[ISI][Medline]

33 Peto R, Darby S, Deo H et al. Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ 2000;321:323–29.[Abstract/Free Full Text]

34 Kuh D, Ben-Shlomo Y (eds). A Life Course Approach to Chronic Disease Epidemiology. Oxford: Oxford University Press, 1997.

35 Naijman JM, Davey Smith G. The embodiment of class-related and health inequalities: Australian policies. Aust NZ J Public Health 1999; 88:216–22.

36 Joseph KS, Kramer MS. Review of the evidence for fetal and early childhood antecedents of adult chronic disease. Epidemiol Rev 1996;2:158–74.

37 Marmot MG, Adelstein AM, Robinson N, Rose GA. Changing social-class distribution of heart disease. BMJ 1978;2:1109–12.[ISI][Medline]

38 Skegg DC. Hormone therapy and heart disease after the menopause. Lancet 2001;358:1196–97.[CrossRef][ISI][Medline]

39 Marmot MG, Bosma H, Hemingway H et al. Contribution of job control and other risk factors to social variations in coronary heart disease incidence. Lancet 1997;350:235–39.[CrossRef][ISI][Medline]

40 Morris RW, Whincup PH, Lampe FC et al. Geographic variation in incidence of coronary heart disease in Britain: the contribution of established risk factors. Heart 2001;86:277–83.[Abstract/Free Full Text]

41 McKee M, Shkolnikov V. Understanding the toll of premature death among men in eastern Europe. BMJ 2001;323:1051–55.[Free Full Text]

42 McMichael T. Human Frontiers, Environments and Disease. Past Patterns, Uncertain Futures. Cambridge: Cambridge University Press, 2001.

43 Beaglehole R. Global cardiovascular disease prevention: time to get serious. Lancet 2001;358:661–63.[CrossRef][ISI][Medline]

44 Rose G. The Strategy of Preventive Medicine. Oxford: Oxford University Press, 1992.

45 Bennett S, Magnus P. Trends in cardiovascular risk factors in Australia. Results from the National Heart Foundation’s Risk Factor Prevalence Study. Med J Aust 1994;161:519–27.[ISI][Medline]

46 MacGregor GA, de Wardener HE. Salt, Diet and Health: Neptune’s Poisoned Chalice: The Origins of High Blood Pressure. Cambridge: Cambridge University Press, 1998.

47 Willett W. Nutritional Epidemiology. 2nd Edn. New York: Oxford University Press, 1998.

48 Egger G, Swinburn B. An ‘ecological’ approach to the obesity pandemic. BMJ 1997;315:477–80.[Free Full Text]

49 Bitton A, Fichtenberg C, Glantz S. Reducing smoking prevalence to 10% in five years. JAMA 2001;286:2733–34.[Free Full Text]

50 McKinlay JB, Marceau LD. To boldly go ... Am J Public Health 2000; 90:25–33.[Abstract/Free Full Text]