Risk factors for disease, as identified by epidemiologists, are properties of individuals which appear to render them more liable to a disease under consideration. These properties may be internal, including genetic characteristics, or external such as indulgence in practices such as smoking or drinking alcohol, exposure to environmental agents or those broader aspects of environment which can be subsumed under the label of socioeconomic status.
It is almost exclusively the case that the rates of occurrence of a chronic disease in human populations are influenced by more than one such risk factor.
Risk factors exert their effect through biological changes in those exposed to them. Such changes tend to subvert normal physiology and biochemistry into pathophysiological pathways which may lead to disease.
There is seldom, if ever, a single such pathway which leads to a particular disease. Rather, a number of different pathways can exist and that which pertains in an individual disease depends upon the particular risk factors pertaining in an individual. The development of a complete pathway to disease is not deterministic but is the result of the combination of those mechanistic components which happen to be present, and whose presence may be dependent upon risk factors. It is possible, however, that the presence of one mechanistic component may have a selective effect upon those others which may be effective.
Risk factors seldom have unique or specific effects but more frequently tend to promote general categories of biological effect, such as genotoxicity, oxidative stress or inflammation. Thus, different risk factors may make incremental contributions to a common form of biological disturbance.
Risk factors need not, therefore be independent of each other in biological effect. Neither need they independent be independent of each other in their occurrence and distribution in a population. Through complex behavioural and social interactions, those exposed to some risk factors may be more liable to be exposed to certain others.
All this being so, attempts to account for rates of disease which emphasise any particular factor, or attempt to allow for the presence of other risk factors by simple statistical correction, are inadequate and misleading.
This website will use information on the relationship between smoking and disease to explore and develop these ideas.
This first edition, however, concentrates on establishing the case that the development of chronic disease is a multifactorial process.
An introductory essay elaborates on the different levels of observation at which risk factors may interact. A simple numerical model demonstrates the possibility that interactions in prevalence between risk factors can affect the estimates of risk which can be derived for any one them. A study using cluster analysis of data from a real population demonstrates that risk factors are not distributed independently in a population but tend to assort or disperse according to the presence or absence of others. The results suggest that smokers tend to carry a greater load of risk factors than do non-smokers.