Industrial atmospheric pollution, historical land use patterns and mortality

Background The measurement of atmospheric pollution for epidemiological studies is problematic. This study presents a new proxy measure of atmospheric pollution of industrial origin and uses it to determine, at electoral ward level, the relationship between atmospheric pollution and all-cause mortality. Methods All-cause Standardized Mortality Ratios (SMR), all ages, and for persons under 65 years for the period 1984-1988, proportions of land in each ward utilized by industrial works (the proxy for atmospheric pollution) and levels of socioeconomic deprivation of the ward residents were compared in 104 electoral wards. Results The all-age SMR in the 22 wards containing the largest proportions of industrial land (113) was 9.7 per cent higher than the SMR (103) in the 60 wards with no industrial land. The under 65 years SMR in the 22 highly industrialized wards (120) was 22.4 per cent higher than the SMR (98) in the wards with no industrial land. After matching the levels of deprivation, the all-age SMR in the 15 wards containing over 10 per cent industrial land (116) was significantly higher than the SMR in 15 wards containing no;industrial land (108); corresponding figures for the under 65 years SMR were 135 and 118. Conclusions A greater proportion of industrial land in a ward is associated with a higher mortality of the ward residents, even after controlling for the level of socioeconomic deprivation of the residents. The association between deprivation and mortality is stronger than the association between atmospheric pollution and mortality. There is an urgent need for better measures of atmospheric pollution which are usable in epidemiological studies.