Primary prevention of cardiovascular disease, with emphasis on pharmacological interventions
Systematic review
|Updated
Key message
We have conducted a systematic review of randomised controlled trials where interventions were implemented for preventing cardiovascular disease among individuals without established cardiovascular disease but who had been diagnosed with hypertension, hyperlipidaemia, diabetes, or otherwise assessed as being at high risk. The quality of the documentation varied considerably.
The results indicate that aspirin may prevent myocardial infarction. However, the quality of the evidence was low, and the potential benefit must be balanced against the risk of bleeding.
We found high quality evidence that antihypertensive and lipid-lowering medication reduces the risk of cardiovascular disease. Among the lipid lowering drugs, the evidence is strongest for statins. Evidence from comparative studies of different antihypertensive drugs did not convincingly show that particular drugs or drug classes were superior. The same can be said for antihypertensive medication for persons with diabetes. Evidence in support of betablockers and alpha-blockers was somewhat weaker than for other drug classes.
For glucose-lowering medication, the results were mixed and the quality of evidence generally low. The results indicate that metformin and acarbose can prevent cardio vascular disease. The results for acarbose were from a study of persons with impaired glucose tolerance. We did not find evidence that sulphonylurea or rosiglitazone reduce the risk of cardiovascular disease.
We identified several studies of multifactorial interventions, usually a combination of lifestyle advice and medication. The results pointed in different directions; it was difficult to conclude whether, or to what extent, such interventions can be expected to have an effect. We included two studies of dietary supplements, one of omega-3-fatty acid, and another of E-vitamin. The results were not convincing with regards to a possible cardiovascular protective effect.
More knowledge is needed about the effects of nonpharmacological interventions.
Summary
Background
The Norwegian Knowledge Centre for the Health Services was requested by the Norwegian Directorate for Health to conduct a systematic review of published scientific literature on the effects of interventions for primary prevention of cardiovascular disease. This was to be used as background material for the development of national clinical practice guidelines in this area.
Methods
We searched for randomised controlled trials in several databases (up to November 2007). The study participants had to include individuals without established cardiovascular disease who had been diagnosed with hypertension, hyperlipidaemia, diabetes, or were otherwise assessed as being at high risk of cardiovascular disease. Studies that did not report on effects in terms of morbidity or mortality were excluded. All references were assessed in terms of relevance, and relevant articles were quality-assessed. When appropriate, the results were pooled in meta-analyses. We graded the quality of the evidence using the GRADE-tool.
Results
The effect-studies we found that fulfilled our criteria for inclusion were mainly about medication. The quality of the evidence varied considerably.
The results indicate that aspirin may reduce the risk of myocardial infarction (22% relative risk-reduction, 95% CI: 3% to 38%) and death (6% relative risk-reduction, 95% CI 0% to 13%). The effect regarding myocardial infarction seems to relate only to men. However, the results also seem to indicate that aspirin protects women, but not men, against stroke (18% relative risk-reduction, 95% CI 4% to 30% among women). The effect of warfarin, or the combination of aspirin and warfarin, has only been evaluated in one study, where only men participated. The results indicate similar effect-sizes as for aspirin alone. The use of anti-thrombotic medication is associated with an increased risk of bleeding, also at low dosages.
Treatment with statins reduces the risk of myocardial infarction (23% relative risk-reduction, 95% CI 18% to 28%). The incidence of stroke also seems to be lower among those who receive these drugs (relative risk-reduction 17%, 95% CI 10% to 24%). This is also the case in regard to total mortality although these findings are less convincing (relative risk-reduction 7%, 95% CI 1% to 14%). For lipid-lowering drugs other than statins the results are less clear, but also these drugs seem to reduce the risk of myocardial infarction (relative risk-reduction 16%, 95% CI 6% to 26%).
Antihypertensive medication reduces the risk of stroke (relative risk-reduction 40%, 95% CI 33% to 46%), myocardial infarction (relative risk-reduction 15%, 95% CI 6% to 23%), development of heart failure (relative risk-reduction 48%, 95% CI 37% to 57%), and death (relative risk-reduction 11%, 95% CI 5% to 16%). Concerning comparisons between different types of drugs or drug combinations, there was no obvious winner. For most comparisons we did not find clear differences, indicating that the most important issue is to reduce blood-pressure per se. In some cases one drug performed somewhat better with regards to one outcome while the other drug performed better for another. Beta-blockers and alpha-blockers were not the statistically significant best agent in any comparison, and the evidence in support of these drug classes can therefore be considered weaker than for the others. The results from the sub-group of diabetes patients largely overlapped with the main findings from these studies.
With regards to the use of antihypertensives as supplementary medication for persons with diabetes, regardless of whether the patient has hypertension or not, the evidence indicates that this has little or no effect on total mortality, risk of stroke or heart failure, but that it may reduce the risk of myocardial infarction (14% relative risk-reduction, 95% CI 1% to 25%) and the risk of developing renal failure (17% relative risk-reduction, 95% CI 4% to 28). We have not found clear evidence that specific antihypertensive drugs stand out as better than others for diabetes patients.
The results from one study of sylphonylurea did not show a clear effect on the incidence of cardiovascular disease. Based on the findings from one study, metformin may have an effect on total mortality (relative risk-reduction 32%, 95% CI 7% to 51%) and risk of myocardial infarction (relative risk-reduction 36%, 95% CI 8% to 55%). It seems likely, based on the results from a single study of persons with impaired glucose tolerance, that acarbose has an effect on the risk of myocardial infarction (relative risk-reduction 92%, 95% CI 36% to 99%). Although the difference was not statistically significant (p>0,05) for most outcomes, the results from the two rosiglitazone-studies we have included provide reason to suspect that the risk of cardiovascular disease increases with the use of this drug.
The results from the studies we have found do not provide a clear answer as to whether, or to what extent, multifactorial interventions are effective for reducing cardiovascular risk. The findings from one study indicate that omega-3 fatty acid (EPA) is not effective with regards to total mortality or risk of stroke, but may be protective against myocardial infarction (22% relative risk-reduction, 95% CI -3% to 41%). Based on the findings from one study, E-vitamin seems to have no effect with regards to preventing cardiovascular disease.
Discussion
The main strength of our work is our systematic approach to retrieving, assessing and summarising the available research-literature. Thus, we believe we have provided a near complete overview of the relevant research. Our assessments of study relevance and quality were partly based on judgements. We have tried to avoid misjudgements by having all assessment done by at least two professionals. Our results are largely in accordance with those found in other systematic reviews in this area.
The reason why we barely address interventions such as diet, smoking-cessation and increased physical activity, is that few studies of such interventions have been conducted where effectiveness was measured in terms of morbidity and mortality.
For lipid-lowering and antihypertensive drugs we found that the results expressed as relative risk-reduction were very similar across included studies. It seems reasonable to assume that women, men, the elderly, and the young have comparable gains from using these drugs, in relative terms. However, even if the effect in terms of relative risk reduction is equal, the absolute gain is dramatically different for groups that are at low risk of disease compared to those at high risk.
Conclusion
Several pharmacological interventions may be effective for the prevention of cardiovascular disease. For multifactorial interventions the evidence is mixed with regards to prevention of cardiovascular disease. The few studies we identified of food supplements did not provide convincing evidence of an effect on mortality or morbidity.
Further research
More knowledge is needed about the effects of non-pharmacological interventions.