[The use of age as a variable in clinical research]

There are a number of ways to study the effect of age as a variable in clinical research. Where survival is concerned, age is the parameter of interest. In this case, age-specific mortality risks, graphs of the logarithms of the age-specific mortality risks set against age (Gompertz-curves), life expectancies, and Kaplan-Meier survival analysis curves are used. Confounding and effect modification by the age of an association between a determinant and outcome can be recognised by evaluating the connection between the different age groups (stratification). Mathematical adjustments for age in linear, logistic, and Cox regression is easy to carry out but their interpretation is often difficult. Adjusting for age is not always necessary.