Is ‘excess’ mortality in Glasgow an artefact of measurement?

Objectives: A previous investigation of Glasgow's excess mortality showed that the (income) deprivation profiles for Glasgow, Liverpool and Manchester were nearly identical. Despite this, premature deaths in Glasgow were found to be more than 30% higher, and all deaths 15% higher, than in the English cities. This study aimed to explore the extent to which Glasgow's higher mortality could be explained by the use of a potentially more sensitive measure of deprivation employed at a suitably small and consistent geographical spatial unit. Study Design: Analyses of mortality based on the creation of a three-city index of deprivation using rates of ‘car/van ownership’ deprivation for small areas (average population size: 1600) in Glasgow, Liverpool and Manchester derived from the census. Methods: Rates of ‘car/van ownership deprivation were calculated for small areas in Glasgow, Liverpool and Manchester. All-cause and cause-specific standardized mortality ratios were calculated for Glasgow relative to Liverpool and Manchester, standardizing for age, gender and deprivation decile. Results: The overall levels of car/van ownership based deprivation in Glasgow, Liverpool and Manchester, in 2001, differed. Glasgow had a higher percentage of its population who did not have access to a car compared with Liverpool and Manchester. All-cause mortality, after adjustment for age, sex and this measure of deprivation, for deaths <65 years were 15% higher and 8% higher for all deaths for males and females respectively. However, this was lower than the excess observed in the previous study. ‘Excess’ mortality was greatest in the working age groups of 15–44 years and 45–64 years, where it was 23% and 15% higher respectively. For deaths at all ages after adjustment, analysis by deprivation decile showed that excess mortality in Glasgow was seen in half the deciles, including four of the five most deprived deciles. However, the greatest excess was seen in comparison of the least deprived neighbourhoods. For premature mortality (deaths under 65 years), the excess was mainly driven by higher mortality in the five most deprived deciles (6–10); again, however, a high excess was seen in comparisons of the least deprived areas. Conclusions: The higher mortality in Glasgow compared to equally income-deprived Liverpool and Manchester cannot be fully explained by a deprivation index based on lack of access to a car or van, but this index does explain more of the excess than income deprivation. Further work to establish better measures of deprivation and to explain this excess are required