It has been said that people consuming large amount of vegetable have lesser chance of cancer because of beta carotene. However this may be due to increase intake of some other factor in diet (eg dietary fiber). This can be explained by?
This is the typical example of a third factor or third variable which is associated both with the exposure and the outcome and can independently cause the outcome, thereby resulting in a mistaken estimate of the cause effect relationship. This factor is known as the confounding factor.
Multifactorial causation: The causal thinking is different when we consider a noncommunicable disease or condition (e.g., CHD) where the aetiology is multifactorial. Two models are presented in Figures 14 and 15 to explain the complex situation.
- In one model (Fig 14), there are alternative causal factors (Factors 1, 2 and 3) each acting independently. This situation is exemplified in lung cancer where more than one aetiological factor (e.g., smoking, air pollution, exposure to asbestos) can produce the disease independently. It is possible as our knowledge of cancer increases; we may discover a common biochemical event at the cellular level that can be produced by each of the factors. The cellular or molecular factor will then be considered necessary as a causal factor .
- In the second model the causal factors act cumulatively to produce disease. This is probably the correct model for many diseases. It is possible that each of the several factors act independently, but when an individual is exposed to 2 or more factors, there may be a synergistic effect.
- From above, it is reasonable to conclude that "one-to-one" relationship in causation is an oversimplification. In biological phenomena, the requirement that "cause" is both "necessary" and "sufficient" condition is not easily reached, because of the existence of multiple factors in disease aetiology. This has created a serious problem to the epidemiologist, who is in search of causes of disease.