Objectives: Four occupation-based measures were used to derive estimates of social position's effect on all-cause mortality among men and women in a large national cohort.
Methods: The National Longitudinal Mortality Study provided information on principal occupation and 9-year follow-up for 229,851 persons aged 25 through 64 years. Cox's proportional hazards model was used to estimate the age-adjusted risk of death relative to six ordinal categories of social position. The Slope Index of Inequality described average change in death rates across categories.
Results: Risk of death was consistently elevated among persons at lower positions in the social hierarchy. Estimates comparing lowest with highest categories varied within a narrow range (1.47-1.92 for men and 1.23-1.55 for women). However, several discrepancies among analyses were noted. The analysis by US census groups revealed nonlinear associations, whereas those using other scales found incremental increases in risk. Effect modification by sex was observed for analyses by two of the four measures. Race/ ethnicity did not modify the underlying association between variables.
Conclusions: Our analysis complements previous findings and supports, with few qualifications, the interchangeability of occupation-based measures of social position in mortality studies. Explanations for why relative risk estimates were modified by sex are offered.