Risks Factors for Vascular Disease

The last forty or so years of medical research have been characterised by advances in the understanding of vascular disease and an enormous increase in the number of factors shown to have a statistical relationship which appears to be causal with the frequency of heart and cerebral vascular disease. Some of these factors hold great promise for prevention, either at an individual level or a population level, or both. A few of the new areas of research will be reviewed briefly. First however, a few comments will be made on the potential of prophylactic intervention.

Reduction of vascular disease through a reduction of risk factor levels: There are several reasons why most strategies for the prevention of disease, and particularly those that are dependent upon changes in diet and life-style, are likely to be of limited benefit.

First: compliance with a prophylactic measure is likely to be poor. In a trial of the effect of reducing mild and moderate hypertension, only 45% of subjects complied with the recommended anti-hypertensive treatment (MRC 1985). Trials of the effect of dietary advice on cholesterol levels have generally shown very little effect - though it is the few trials which do show an effect that are the most often quoted! (Ravenov 1992). Attempts to change the intake of dietary fibre, or fruit and vegetable intake have achieved very little (Burr et al 1989). Results such as these have profound implications for prophylactic strategies in symptomless subjects, or in the general population.

Secondly: effectiveness is likely to be low. Rose defined a paradox: population preventive strategies, even those with large overall potential benefit, are likely to confer little benefit on the individual, and he went on to describe the overall benefit/risk balance in strategies involving the general population as `worrisome' (Rose 1989). An overview of nine major population based trials of the effect of advice on smoking, diet and blood pressure led to the conclusion that: `The pooled effect of intervention on multiple risk-factors is insignificant'. (Ebrahim & Davey-Smith 1997).

Third: costs are likely to be high. Estimates of cost of saving of a life by population screening and appropriate drug prophylaxis for raised cholesterol levels, is likely to be around £60,000 ($85,000) per year of added life (Ferner & Neil 1995 and others). The prophylactic treatment of `mild' or `moderate' hypertension requires the identification and treatment of 850 to give the possibility of preventing one stroke each year, one coronary event every five years and a death every twenty years (MRC 1985). This, followed through, implies enormous costs for each disease event prevented.

 The `classical' risk factors: smoking, blood pressure and cholesterol:
Cigarette smoking is probably the most clearly established of all the possible risk factors for heart disease. It is directly responsible for a high proportion of heart disease and in some way it is causal in at least 20% of all deaths in the UK. Smoking also appears to act synergistically with other risk factors. There is a rapid decline in the incidence of vascular disease in men who quit smoking, though the delay before the risk returns to that of subjects who had never smoked is likely to be ten years or more.

One of the underlying mechanisms in smoking is a rise in acute-phase proteins including fibrinogen, plasma viscosity and white cell count, each of which is strongly predictive of ischaemic heart disease events. There is a slow decline in the levels of acute phase reactants in ex-smokers, but it takes around ten years for the levels to reach those of men who had never smoked (Yarnell et al. 1987).

Blood pressure has consistently been found to be a major predictive factor for coronary heart disease and stroke and an overview of randomised controlled trials of blood pressure lowering by drugs has shown reductions of around 14% in the incidence in coronary events and about 40% in stroke (Collins et al. 1990).

An early provocative paper on cholesterol, entitled: `Coronary heart disease is not preventable by population intervention' (McCormick & Scrabnek 1988) was followed by a number of supporting overviews of evidence available at that time on cholesterol and heart disease. The basic claim implied in that title has never been refuted and the more recent overview referred to already confirms the almost total ineffectiveness of standard health education methods aimed at cholesterol and the other classical risk factors (Ebrahim & Davey-Smith et al 1997).

Cholesterol lowering has of course entered a new phase with the development of the statin drugs as these have, for the first time, given evidence that cholesterol lowering can affect mortality _ though at a price! (Scandanavian Simvastatin Survival Study Group 1994). An even more fruitful line of research, however, and one that is likely to lead to more acceptable and less expensive prophylaxis, arises from the observation that LDL cholesterol has to be oxidised before it becomes atherogenic. Much research is therefore focussing on dietary fruit and vegetable intakes, antioxidant vitamins and enzymes and on oxidised LDL itself. (O'Keefe et al 1995). Another intriguing finding is that statins do more than just lower cholesterol (Vaughan et al 1996). They have effects on haemostatic and other mechanisms and the very early benefit seen in trials has been attributed to effects they have on coagulation (Shepherd 1995).

Early nutrition:
One of the most intriguing recent observation on heart disease, is a relationship between the disease in later adult life and indices of nutrition early in life, the so-called Barker hypothesis (Barker 1994; Frankel et al 1996). Thus birth weight, which is a resultant of both genetic and intrauterine influences, shows a strong association with later risks for both heart disease and diabetes. In one study of a representative cohort of men in South Wales, the fifth of men who had had the heaviest birth weights had a later incidence of heart disease which was one third that of the fifth of men with the lightest birth weights (Frankel et al 1996). A strong relationship is also shown with the height of adults, which is again an outcome of both genetic influences and nutritional factors during childhood and adolescence. In the same cohort of men, the difference in heart disease risk was two-fold in favour of the tallest men. These relationships are robust and do not appear to be explained by general socioeconomic factors _ either those which pertained during childhood, or those relevant to the subjects when adult.

Chronic infections:
A number of cross-sectional and case-control studies have reported positive associations between ischaemic heart disease and infections with various chronic infecting organisms including Chlamydia pneumonia and Helicobacter pylori (Strachan et al 1998; Mendall et al 1994). While these organisms are more usually associated with mild respiratory infections and chronic gastritis, a hypothesis has been suggested to explain a possible association with heart disease. Infection leads to a rise in factors predictive of ischaemic heart disease, including C-reactive protein, fibrinogen and white cell count.

It would be of great interest if the association with chronic infections were confirmed, both because part, and possibly a substantial part, of the social class gradient in ischaemic heart disease and stroke might be explained, and because the chronic infective agents which are most suspected, can be easily eradicated. The evidence has however been reviewed recently (Danesh et al 1997) and judged unconvincing.

Homocysteine:
In a rare inherited disorder there is impaired clearance of homocysteine, a metabolite of the amino acid tryptophane. This leads to a huge elevation in serum levels and affected patients show markedly accelerated atherosclerosis and thrombosis (Rees and Rogers 1993). This observation led to the hypothesis that elevations of homocysteine which are within a range that would be considered as normal in clinical practice, may carry an increased risk of coronary, cerebral and peripheral atherosclerosis (Bousey et al 1995). However inconsistencies in the evidence from case-control  and the more powerful prospective studies indicate the need for evidence, and in particular, evidence on the effect of folate supplementation on the incidence and progression of vascular disease (Clarke and Collins 1998).

Alcohol:
Alcohol is responsible for a considerable burden of morbidity, mortality and social disruption. Nevertheless, moderate alcohol intake has been found to be negatively associated with ischaemic heart disease mortality in a number of studies, the incidence in drinkers being between 30 and 60% lower than in men who drink only occasionally, or not at all. (Doll et al 1994; Poikolainen 1995). At the same time, the balance of the effects of alcohol on blood pressure, and on mortality from heart disease is likely to be unfavourable in the heaviest drinkers (Hart et al 1999).

The mechanism which has been most often suggested to explain the beneficial associations in vascular disease is an increase in high-density lipoprotein cholesterol in drinkers. Even after allowing for this effect however, there is still a marked protective effect of alcohol. This residual effect, is probably due to the action of alcohol on haemostasis. Alcohol has favourable influences on a large number of haemostatic factors, including a substantial reduction in the response of platelets to several agonists (Renaud et al 1992). This last led Renaud & deLorgeril (1992), commenting on the remarkably low IHD mortality in France - the so called French paradox _ to suggest that protection is because the wine which is widely consumed in France, substantially reduces platelet aggregation.

Magnesium:
Magnesium is an essential element and low tissue levels, such as occur in alcohol intoxication, are associated with cardiac arrythmias and death. Magnesium has many physiological effects (Woods 1992) and a number of studies have found that levels in the myocardium are substantially lower in subjects who have died following a heart attack compared to levels in subjects who have died from other causes (Chipperfield & Chipperfield 1973; Elwood et al 1980.

These and other findings led to the practice of giving intravenous magnesium early in the acute phase of myocardial infarction. Conflicting results from two large trials have however thrown this into confusion. LIMIT-2 showed a highly significant reduction in clinical outcomes in patients given an infusion of magnesium soon after infarction (Woods et al 1992) while ISIS-4 gave no evidence of protection by magnesium (ISIS-4 1995).

An even more important question is whether or not chronic magnesium deficiency, of a degree sufficient to impair myocardial function occurs in the community, and whether or not a magnesium supplement would be beneficial. Evidence from trials is required.

Haemostastic factors:
The huge rise in cardiovascular mortality during the first seventy years of this century, has been followed in more recent years by an equally rapid fall. What evidence there is suggests that neither the prevalence nor the severity of atherosclerosis of the vessel wall has changed much during this period. On the other hand, infarction of the myocardium appears to have been rare at the beginning of this century. Although there can be no certainty, all this suggests that changes in the incidence of cardiovascular disease have resulted from changes in factors involved in thrombosis, rather than due to any marked changes in the prevalence or severity of arterial atherosclerosis.

While it is virtually impossible now to gain evidence about changes over time, it has been conclusively shown that fibrinogen, plasma viscosity, von-Willebrands factor and other factors involved in thrombosis are strongly predictive of thrombotic events such as coronary artery thrombosis (Yarnell 1991). Work on these factors is required to identify their determinants, that is, the dietary and life-style factors which determine their level, and hence the risk of a thrombotic event (Elwood et al 1993). The most powerful determinant which has been identified so far is smoking (Yarnel et al. 1987). 

Blood platelets:
Platelets play a key role in thrombosis and they constitute the link between atheroma and a occlusive thrombosus in mycardial infarction and stroke.

Relevant evidence comes from a wide variety of sources, including the histology of thrombi in the corony vessels (Davies & Thomas 1984), the finding of platelet emboli in the micro circulation of the myocardium after sudden death (Haerem 1974) and the finding of an occlusive thrombus on angiography shortly after infarction (Dewood et al 1980). Several case-control studies have shown that platelet aggregation is enhanced after myocardial infarction (Dreyfus & Zahave 1973) and cross-sectional evidence shows an association between aggregation and prevalent IHD (Elwood et al 1990; 1991).

The strongest evidence that platelets are a key factor however, is the effect of aspirin and other platelet active drugs on the risk of cardiovascular disease events. This is the main topic of the rest of this booklet.

On the other hand, there is no convincing evidence that any measure of platelets can be used in prediction, that is, to identify subjects who are at increased risk of a thrombotic event. Although two small studies were suggestive, a major long-term cohort study gave no evidence of any prediction for heart disease by the present conventional tests of platelet aggregation (Elwood et al 1998). A measure of platelet activity which is predictive of IHD events could be of very great value in clinical practice. If such a test could be developed it could be used as a screening procedure to enable low-dose aspirin prophylaxis to be targeted on patients found to have more highly reactive platelets.