Objective: To examine associations between area and individual socio-economic characteristics and premature cancer mortality using multilevel analysis.
Methods: We modeled cancer mortality among 25-64-year-old men and women (n = 16,340) between 1998 and 2000 in Australia. Socio-economic characteristics of Statistical Local Areas (n = 1,317) were measured using an Index of Relative Socio-economic Disadvantage (quintiles), and individual socio-economic position was measured by occupation (professionals, white and blue collar).
Results: After adjustment for within-area variation in age and occupation, the probability of premature cancer mortality was highest in the most disadvantaged areas for all-cancer mortality for men (RR 1.48 95% CI 1.35-1.63) and women (RR 1.30 95% CI 1.18-1.43) and for lung cancer mortality for men (1.91 95% CI 1.63-2.25) and women (1.51 95% CI 1.04-2.18). Men in blue collar occupations had a higher rate of cancer mortality (RR 1.57 95% CI 1.50-1.65) and lung cancer mortality (RR 2.31 95 % CI 2.09-2.56), whereas men in white collar occupations had a lower all-cancer mortality rate (RR 0.78 95% CI 0.72-0.85). Compared with professionals, women in white collar occupations had an all-cancer mortality rate that was lower (RR 0.85 95% CI 0.80-0.90). When deaths from breast cancer were excluded, women in blue collar occupations had a significantly higher all-cancer mortality rate than professionals (RR 1.12 95% CI 1.02-1.22).
Conclusions: Area disadvantage and individual socio-economic position were independently associated with premature cancer mortality, suggesting that interventions to reduce inequalities should focus on places and people.