The Use of Age‐Specific Mean Cohort Slopes in the Analysis of Epidemiological Incidence and Mortality Data

In studies of disease causation, analyses of incidence and mortality rates are frequently used to identify sex—or population—specific variations which may be related to aetiologic mechanisms. The relationship of the first differences of the rates, averaged over cohorts, with age has proved useful in the detection of relatively small variations, and has yielded inferences not obvious from the rates themselves. This relationship is characterized here by a polynomial regression equation, the parameters of which are estimated by a maximum likelihood method which uses a general variance—covariance matrix for the residuals, and which takes account of the particular shape of data matrix which arises in studies involving vital records. An example of the calculations is given for some cancer incidence data.