The structure of urban mortality. A methodological study of Hannover, Germany. Part II

An extension is made of a prior analysis of the relation of density to cause-specific mortality in small areas of Hannover, Germany. The analysis was designed to impose statistical controls for certain socio-demographic characteristics of these small areas under a model where all of the effects of density were assumed to operate on mortality only indrectly by affecting the residential decisions of select socio-demographic groups. It was found that, despite the sizeable effects of socio-demographic factors on cause-specific mortality, density retained a direct effect on five of 14 cause-specific rates. Implications of these results for modelling density-mortality relations are discussed.