Using the Environment to Explain and Predict Mortality

Mortality at ages 45-74 has been studied in the larger County Boroughs of England and Wales. In the younger segment of this span of years premature death (and morbidity) from the chronic diseases is now making large inroads into our manpower. Although this study does not look at inherited predispositions or individual behaviour habits, the socio-environmental patterns are described for the major causes of death. Large differences in death rates still persist between towns in this country, and are seen to be related to changing, changeable and unchangeable aspects of their environment. Multiple regression methods form the basis of the analysis of the large data bank collected for this purpose. The socio-environmental variables are shown to be differently related to different causes of death, and for some causes a very high proportion of the variance of death rates between towns is explained. An examination of the effectiveness of these equations in predicting the mortality in a further set of towns not included in the initial analysis is undertaken. This is looked at in relation to a theoretical result showing how well an estimated regression line can be expected to perform in a second, independent set of data. Although the multiple regression model is almost certainly not correct in this situation, it is interesting how closely the results fit. A further section looks at the problems of interpreting the regression coefficients when the regressor variables are highly intercorrelated, and gives some guidance in the particular situation.