Why is mortality higher in poorer areas and in more northern areas of England and Wales?

STUDY OBJECTIVE

To identify and quantify the factors responsible for the differences in mortality between affluent and deprived areas, the north and the south, and urban and rural areas in England and Wales.

DESIGN

A multiple Poisson regression analysis of cause specific mortality in the 403 local authority districts, each classified by deprivation (using the Jarman Index), latitude (from 50 degrees to 55 degrees north) and urbanisation, adjusting for age, sex, and proportion of ethnic minorities.

SETTING

England and Wales 1992.

MAIN RESULTS

All cause mortality was 15% higher in the districts comprising the most compared with the least deprived tenth of the population, 23% higher in the most northern (55 degrees) than in the most southern (50 degrees) districts, and 4% higher in metropolitan (within large cities) than rural districts. Nationally these differences were associated with 40,000, 65,000, and 15,000 excess deaths respectively. More than two thirds of the overall excess mortality with deprivation, latitude, and urbanisation was from three diseases--ischaemic heart disease, lung cancer, and chronic bronchitis and emphysema. The excess mortality from these and other diseases closely matched that predicted from differences according to deprivation and latitude in smoking, heavy alcohol consumption, Helicobacter pylori infection, and temperature, and thus could be attributed to these causes. About 85% of the overall excess mortality with deprivation was attributable to heavier smoking and 6% to heavier alcohol consumption, but diet varied little. Deaths more directly related to deprivation (such as those caused by H pylori infection, drug misuse, psychoses) accounted for an estimated 12% of the excess deaths, but variation in provision and uptake of healthcare services only 1%. The direct effects of deprivation are more strongly related to morbidity than mortality. Of the difference in mortality with latitude, about 45% was attributable to differences in smoking, and 25% to climate (mainly the association of cardiovascular and respiratory disease with cold). The differences with urbanisation were mainly because of smoking.

CONCLUSIONS

Differences in the prevalence of smoking account for much of the variation in mortality between areas. Alcohol accounts for some, diet little. The more direct material effect of deprivation contributes to the variation in mortality but is particularly important with respect to differences in morbidity.