Studies on the prevention of a cardiovascular disease in a population carry specific scientific problems, as recently summarized by Thelle . In individuals (high-risk approach) risk factors can be identified and intervened upon by non-pharmacological and/or pharmacological tools. The changes in risk factor levels can easily be monitored. It becomes much more complex to determine attitudes, behaviours, lifestyle changes, the risk factor burden and outcome measures such as mortality and morbidity in the whole population. The cause/effect relationship is thus harder to identify. Furthermore, the risk in healthy ‘normal’ individuals is much lower than in high-risk individuals. Clinical events will thus occur much more infrequently in such normal populations than in high-risk groups, necessitating much larger study groups to achieve sufficient statistical power. It is therefore not astonishing that the population-based intervention programmes have had problems in showing positive results [6–9]. The only other Swedish project that is comparable to the MPP is the Gothenburg Primary Preventive Trial (GPPT) . In this study, middle-aged men were screened and treated for cardiovascular risk factors, but no overall beneficial effect could be detected in total or CVD mortality in comparison with a predetermined control group . Dietary advice, smoking cessation and drug treatment, if necessary, for hypertension and hyperlipidaemia were all integrated parts of this programme, but there were no activities focusing on alcohol overconsumption or health problems of women such as those in the MPP. The similarities between the two studies are obvious and neither of them could show a preventive effect on cardiovascular outcomes over and above that seen in the general background population of the two cities.
Surprisingly, the WHO, the National Institute for Heart, Lung and Blood Disease (NHBLI), as well as other organizations have all recommended radical changes in lifestyle in spite of the lack of scientific evidence for a preventive effect of large-scale community studies. Included in the lifestyle advice are measures such as smoking cessation, improved diet, decreased alcohol consumption and enhanced physical activity for whole populations. Other lines of argument must have been used when these political health goals were set. Apart from cessation of smoking, the suggested lifestyle changes have a relatively weak scientific basis for effectiveness on the population level. The Swedish Council of Technology Assessment in Health Care has recently published an extensive review of population-based preventive efforts in defined communities . This review identified eight large preventive projects in Europe and the USA with the whole population as target group. For inclusion, these projects should:
be directed towards more than one risk factor;
follow the changes in risk factors, morbidity and mortality in a comparable control population as well;
have published the results in scientific journals.
In summary, these large community projects showed no or insignificant beneficial effects in the intervention communities as compared with the control areas , and the review presented the following conclusions and recommendations.
Smoking, high blood pressure, unhealthy dietary habits, low physical activity and social factors seem to play a role in the development of CVD. However, the review gave no scientific basis for starting new, large population-based intervention projects with the tools that were used in the scrutinized projects.
Several attempts to prevent CVD are in progress in Sweden and elsewhere. They are seldom designed in a way that makes proper evaluation of the effects possible. The scientific competence in these projects should therefore be strengthened.
The evaluation of the intervention effects in population-based projects needs development of new methodology, not least within the social and behavioural processes involved in preventive efforts in the population.
The review deals with preventive projects with design and methods of the 1970s and 1980s, with the inherent problem of generalizing the findings to the present situation. The important task of preventing disease should be based on scientifically proven facts. The knowledge gained from these large population-based preventive projects should be looked upon as a basis for more focused studies on how CVD can and should ideally be prevented in the future.
Why did the large projects fail to show expected results? There are several explanations. The similar risk factor decrease in both intervention and control areas in most studies, which might imply seeding of information from the project area to the control area, is one major possibility. Another possibility is use of ineffective methods for diagnosis, treatment and follow-up with regard to both lifestyle and pharmacological interventions. A third is lack of knowledge of the psychological and social mechanisms involved in changing attitudes and behaviours, not only of individuals but of whole populations. In future preventive programmes, social scientists should therefore ideally be involved in both design and implementation at an early stage. Whatever the reason(s), the findings of the large population-based project call for future research with more stringent design and more effective interventions .
High-risk individual approach
The three well-known major risk factors for CVD – hypertension, hyperlipidaemia and smoking – have all been addressed in randomized controlled trials (RCTs). Different populations and treatment strategies have been involved, but mostly with a selection bias towards middle-aged Caucasian men and drug treatment. The overall conclusion to be drawn from a meta-analysis of antihypertensive RCTs is that a mean net diastolic blood pressure reduction of about 5 mmHg has been associated with a 40% relative reduction of stroke and a 14–18% relative reduction in the incidence of myocardial infarction , having the largest benefits in elderly hypertensives. Older antihypertensive drugs are equally as effective as more modern ones in elderly hypertensives . The angiotensin-converting enzyme (ACE) inhibitors have additional protective properties independent of the blood pressure effect, as recently shown with ramipril in the HOPE study . It should, however, be pointed out that in these cited studies, the absolute risk reduction has been rather modest, being less than 10% over a 3–5-year time period. The absolute risk is the only risk of personal importance to the patient.
Treatment of hyperlipidaemia has been well documented in the 1990s for both primary prevention [30, 31] and secondary prevention [32–34], and seems cost-effective in specific subgroups of patients, e.g. in type 2 diabetes in a secondary preventive trial . Statins have mostly been used in recent RCTs and are proven to be safe and effective [30–34]. However, the absolute risk reduction in these studies has been modest. In Malmö, the treatment of hyperlipidaemia was an integral part of the intervention programme of MPP, focusing on the special risk category of combined hypercholesterolaemia and hypertriglyceridaemia, with good results on risk factor levels [16, 17].
Smoking is perhaps the risk factor for CVD which theoretically should be the most cost-effective to treat, and is a very well proven risk factor for many disorders in long-term observational studies. Advice for smoking cessation has been incorporated more or less successfully in the majority of RCTs for prevention of CVD, but it should be pointed out that well-designed intervention studies on smoking cessation are few. One of the most convincing studies so far to prove the benefits of smoking cessation is the Oslo Diet and Anti-smoking Study with very favourable long-term results . A more healthy diet in combination with smoking cessation was shown to decrease CVD mortality drastically. In the MPP, smoking cessation was recommended to many patients at risk, but evaluation of separate effects of quitting smoking has not been possible due to insufficient follow-up data on smoking habits.