Age, socioeconomic status, and mortality at the aggregate level.

STUDY OBJECTIVE Indicators of socioeconomic status are associated with age. This study aimed to analyse the influence of the age distribution on the ranking of small areas by socioeconomic status and on the association between their socioeconomic status and standardised mortality. DESIGN The ranking of small areas by socioeconomic status indicators (educational level, income, and unemployment) was compared with crude values and after correction for their age structure. The age and gender standardised mortality ratios (SMRs) of these areas for the age group 1-64 years was then rank correlated with both crude and age standardised measures of socioeconomic status. SETTING This study used data for all (n = 22) boroughs of Amsterdam for the period 1986-91. MAIN RESULTS Correction of indicators of socioeconomic status for the age structure of the population hardly affects the ranking of Amsterdam boroughs by socioeconomic status. All rank correlations between crude and age standardised socioeconomic status measures are above 0.95. Rank correlations between SMR and these socioeconomic status measures also hardly change after correction for the age structure of boroughs except for education. Mean income per earner is the socioeconomic status indicator most strongly associated with the SMR. CONCLUSIONS This study shows that the age structure of Amsterdam boroughs has almost no influence on their ranking by socioeconomic status and a limited influence on the association between their socioeconomic status and SMR, except for educational level. The latter indicator has the strongest association with age. This result and theoretical considerations indicate that a correction for the age structure of the population will be more important if small areas differ little with regard to socioeconomic status, if they vary considerably in age structure, or if a given indicator of socioeconomic status shows a strong cohort effect or age association.