Abstract

Objectives: To measure geographical co-relationships between disease-specific standardised mortality ratios (SMR) and different atmospheric emissions in 352 English local authorities. To link specific exposures with specific causes of death and to identify responsible polluting sources. To see whether long-term moderate exposures have the same lethal effects as short-term high-pollution (ie, smog) episodes.

Design: Geographical distributions of SMR, atmospheric emissions and social hazards, extracted from three different sources, were converted to a congruent format. Correlation coefficients were calculated within and between these different datasets. Mortality/pollutant correlations were recalculated after additionally standardising the SMR for social differences between local authorities.

Setting: The 352 English local authority areas, 1996–2004.

Main results: SMR for one group of diseases (including upper alimentary and respiratory cancers, ischaemic heart disease, peptic ulcer, pneumonia) were related to a range of combustion emissions and to multiple social deprivation, cigarette smoking, binge drinking and a northern location. Additional standardisation of all SMR for these social hazards left a small subgroup independently related to atmospheric pollution, mainly from oil combustion. Correlations with pneumonia deaths were exceptional.

Conclusions: High mortality rates were observed in areas with elevated ambient pollution levels. The strongest single effect was an increase in pneumonia deaths. Road transport was the chief source of the emissions responsible, although it was not possible to discriminate between the different chemical components. Many “pneumonia” deaths were probably caused by direct chemical injury, as in the 1952 London smog and are better regarded as “acute respiratory distress syndrome” or “acute lung injury”.