Search

The relation between mortality of people aged 45 to 74 and atmospheric smoke and sulphur dioxide in the county and London boroughs of England and Wales was investigated using data for the years 1969 to 1973. Weighted multiple regression analyses, which took into account a number of socioeconomic and climatic variables, showed no consistent relation of smoke or sulphur dioxide with mortality from all causes or with mortality from specified causes postulated a priori to be related to pollution. In particular there was no significant association between smoke and mortality rates for respiratory illness. Comparison with results from similar analyses of data for the two previous decades suggested that a decline in the strength of associations had occurred in parallel with declining levels of the pollutants.

A study was set up to investigate the effects of annual changes in the levels of atmospheric smoke and SO2 on changes in health from 1973 to 1977 in primary schoolchildren from 28 randomly selected areas of England and Scotland. Changes in health were measured by taking the change in number of respiratory conditions reported from one annual examination to the next. The number of areas with data on pollution in each period was 5,9,17, and 14 respectively and within these areas the response rate varied from 65% to 74%. Altogether 857, 1436, 2702, and 2036 children respectively who were of white ethnic origin, aged 6 to 11, and had complete data on sex, social class, and changes in health were studied in each period. In 1973-4 the levels of pollution were highest and showed the greatest decline. The greatest annual mean change in smoke was a decrease from 71.9 to 50.5 microgram/m3 and in SO2 a decrease from 94.2 to 47.6 microgram/m3. However, no relation was found between improvement in health and decreasing levels of pollution. In subsequent years, when the levels of pollution were lower and showed smaller changes, change in health was also unrelated to changes in pollution. Thus no evidence was found to suggest that the levels measured during the study were harmful to health.

The relation between respiratory illness and atmospheric smoke and sulphur dioxide (SO2) was investigated from 1973 to 1977 in children aged 6 to 11 from a random sample of 28 areas in England and Scotland. Cross-sectional results are presented for 1975, and results from other years briefly summarised. In 1975 there were 19 areas with data on pollution and in these areas the sample included 5787 children of white ethnic origin of whom 4116 (71%) had complete information of respiratory illness and other variables considered in the analysis. After allowing for the effects of age, social class, population density, type of fuel used for cooking in the home, and season of examination, the prevalence of respiratory illness in both sexes was in the home, and season of examination, the prevalence of respiratory illness in both sexes was positively associated with the levels of smoke over the range of annual means 8 to 51 microgram/m3 )P less than or equal to 0.05). No relation was found between illness and annual means of SO2 ranging from 12 to 114 microgram/m3. Similar results were found in other years, and in 1977, when information of tobacco smoking at home was collected, the association between illness and atmospheric smoke appeared to be independent of smoking within the home. The levels of smoke were much lower than those at which effects on health hve previously been reported so the association is unlikely to be causative. We postulate that higher levels of atmospheric pollution at an earlier period in some areas may have predisposed children living there to respiratory illness during their primary school years. Alternatively, some other characteristics of the polluted areas may explain the findings.