4. Measures of effect
Epidemiologic research aims to measure effects by comparing risks or rates.
We measure effects by comparing risks or rates in two or more groups. If smokers have a rate of lung cancer that is 10 times the rate among non-smokers, this is a measure of the effect of smoking on lung cancer.
We can compare either rates or risks by looking at the absolute difference between the rates or risks, or by looking at their relative size (comparison made in the terms one rate or risk over another, e.g. twice as big).
Example: Suppose the risk of a smoker getting lung cancer between 30 and 70 years of age is 8%, and the risk of a non-smoker getting lung cancer between those ages is 1%. The risk difference (absolute effect) over the 40 years is 7%, i.e., by smoking you are increasing your chance of getting lung cancer by 7%. The risk ratio (relative effect) over the same period is eight, smokers are eight times more likely to get lung cancer than non-smokers in this age range.