The Multifactorial Aetiology of Coronary Heart Disease: A Dangerous Delusion

THE MULTIFACTORIAL AETIOLOGY OF CORONARY HEART DISEASE: A DANGEROUS DELUSION JAMES McCORMICK* Introduction The term multifactorial when applied to aetiology is a tautology that has led to unreal expectations and delusion. AU diseases are multifactorial in origin. Infectious disease does not inevitably follow exposure to pathogenic organisms; many other conditions need to be satisfied before disease ensues. Road accidents depend on the conjunction of many factors , factors as various as blood alcohol, temper and temperament, eyesight and weather. Nobody refers to road accidents as having a multifactorial aetiology, the phrase being reserved for diseases whose aetiology is unknown, most commonly coronary heart disease and sometimes cancers . The phrase is a synonym for "unknown" and thus a euphemism for ignorance. Risk Factors Are Risk Markers Since the end of the Second World War the new epidemiology has been largely concerned with the identification of risk factors. A risk factor strictly defined is a factor which is associated with an altered probability of disease occurrence. Unfortunately, not everyone makes a clear distinction between associations and causal relationships, and it is common to find doctors and others talking about risk factors as if causal relationships had been established. Such an assumption leads almost inevitably to the belief that modification of risk factors will lead to a reduction of disease incidence. It is this faulty logic which is responsible for the dangerous delusion. It would be better if the term risk factor were dropped and replaced by risk marker. *Department of Community Health, University of Dublin, 196 Pearse Street, Dublin 2, Ireland.© 1988 by The University of Chicago. AU rights reserved. 0031-5982/89/3201-0613$01 .00 Perspectives in Biology and Medicine, 32, 1 ¦ Autumn 1988 | 103 The initial identification of risk markers usually derives from casecontrol studies, but associations have also been described in prospective studies. Case-control studies are a quick, dirty, and relatively cheap way of generating aetiological hypotheses, and their use has been extremely fruitful. Nonetheless, they are prone to bias and never of themselves prove causal relationships [I]. The results of such studies are usually expressed in terms of odds ratio or relative risk, nowadays happily accompanied with confidence limits. As a rule the relative risks which emerge from most of these studies are small, usually less than two. Quite inappropriately, such relative risks once published are often taken up by the popular press and some enthusiasts and translated into advice about life-style or the avoidance of this pill or that article ofdiet. Because ofthe flaws in the method and the possibility of error, small relative risks so derived should never serve as a basis for action other than further, preferably prospective or experimental, studies. Good prospective studies provide more secure evidence of the strength ofassociations, but only experiment can prove causality. The list of risk "factors" for coronary heart disease grows daily. It now includes, and this is by no means an exhaustive list, cigarette smoking, serum cholesterol, hypertension, obesity, diabetes mellitus, low levels of high-density lipoproteins, high levels of low-density lipoproteins and triglycerides, selenium, thiazide diuretics, not drinking, not exercising, not having siestas, not eating fish, especially mackerel, living in Scotland, having a high level of phobic anxiety, being scrupulous about keeping appointments, not taking cod-liver oil, and snoring. The important associations include age, being male, a family history, a high level of fibrinogen , and—perhaps most important of all—being poor in the rich world. Multifactorial Aetiology On the basis of such associations it is stated that diseases such as ischaemic heart disease have a multifactorial aetiology. That is, unlike infectious disease and accidents, they do not have a single necessary cause but are the result of some combination or amalgam of risk markers , none of which is by itself necessary. The concept of cause is difficult. Even simple examples, such as being hit on the head by a falling hammer, can be made complex. The immediate cause, that is, the falling hammer, is both sufficient and necessary , but the antecedent causes involve being under the scaffolding at the right time and can be pursued backward in time to the accident of birth without which the possibility ofbeing rendered unconscious in...