Measures of effect in epidemiological research.

The study of the relationship between risk factors and outcomes is important both in etiological and prognostic research. To assess the strength of a given risk factor-outcome relationship we use measures that are calculated in relative and absolute terms. Risk ratio, incidence rate ratio and odds ratio are relative measures of this relationship. Risk difference (or attributable risk) and rate difference (or attributable rate) are absolute measures of the same relationship. Risk difference and rate difference are calculated by subtracting the risk and the incidence rate in exposed individuals from that in unexposed individuals, respectively. The choice of these measures depends on the study aim. Relative measures are commonly used in etiological studies while absolute measures are mainly used in public health research.