Proving Causality

What caused my cancer? This is one of the most common questions people ask about cancer and there is no simple answer. The following are some points of information that need to be considered.

Association versus cause An association between an exposure and cancer DOES NOT necessarily mean that the exposure caused the disease.1 A common perception is that if several persons in a geographic area develop cancer and there was once a chemical manufacturer, or landfill, or some industry in the area, then that is what caused the cancers.

Multiple causation or multi-factorial etiology It is a common perception that cancer is a single disease. However, cancer is actually many different diseases, each with differing rates of occurrence, risks, causes, and chances of survival. Not all persons develop the same disease for the same reason (i.e., no one factor determines whether an individual will develop a disease). It is the interaction of many factors that produce disease (e.g., for cancer this could be genetics, immunity, diet, occupation, hormonal, viral, socioeconomic, lifestyle, or physical environment).1

Can a study be done to prove what caused my cancer? Finding causes of any disease is usually a long, slow process. No one study is likely to prove that a particular exposure definitely causes a particular disease. No single study, or even a large number of epidemiologic studies, will enable a person to know why he/she developed cancer.1

Scientific evidence needed to establish cause and effect between exposure and occurrence of disease.1-5

Consistency Consistency is often referred to as reproducibility. The association between the exposure and the disease is observed in different studies, among different populations, under different settings and using different study methods.

Strength There is a reasonably high relative risk or odds ratio associated with the exposure or frequency in which the exposure is found in the disease (e.g., a ten- fold risk is stronger than a two-fold risk). The frequency with which the exposure occurs in the absence of disease must also be considered.

Dose-relationship Dose-relationship is between the level of exposure and disease, or the risk of the disease increases or decreases as the amount of the exposure increases or decreases.

Biologically plausible The relationship between the exposure and the disease is based on proven biological or medical knowledge (i.e., more confidence in an association if it has already been shown to cause disease in animals).

Temporally correct There is a reasonable time between exposure and occurrence of disease (latency period), and that the exposure preceded the development of disease. Specific cancers may vary somewhat in the length of the latent period, but generally speaking, recent exposure, that is exposures in the last 10 years, cannot be expected to be associated with current cancer incidence.

References

1. Ries LA, Smith MA, Gurney JG, Lind M, Tamra T, Young JL, and Bunin GR (eds.) Cancer Incidence and Survival among Children and Adolescents: United States SEER Program 1975-1995, pp. 9-12, National Cancer Institute, SEER Program. NIH Pub. No. 99-4649. Bethesda, MD. September 1999.

2. Nachman Brautbar, MD, Chemicals and Cancer: Establishing Causation through Medical Toxicology. Article retrieved October 2005 from website: http://www.environmentaldiseases.com/article_chemicals_cancer.html

3. Environmental Health Criteria 72, Principles of Studies on Diseases of Suspected Chemical Etiology and Their Prevention. World Health Organization, Geneva, 1987.

4. Rom WN. Causation, in Environmental and Occupational Medicine (2nd Ed), 1992. Little Brown and Co., Boston, Toronto and London.

5. Rothman KJ, Pool C. Causation and causal inference. In Schottenfeld D, Fraumeni JF (Eds). Cancer Epidemiology and Prevention. Philadelphia: WB Saunders Co; 1996:1-10.