Objectives: (i) examine the relation, if any, of pre-morbid IQ scores at 20 years of age with the risk of later cancer mortality; and (ii) explore the role, if any, of potential mediating factors (e.g. smoking, obesity), assessed in middle age, in explaining the IQ-cancer relation.
Methods: Cohort study of 14, 491 male, Vietnam-era, former US army personnel with IQ test scores at around 20 years of age (1965-71), who participated in a risk factor survey at around age 38 years of age (1985-6), who were then followed up for mortality experience for 15 years.
Results: There were 176 cancer deaths during mortality surveillance. We found an inverse association of IQ with later mortality from all cancers combined (age-adjusted HR(per one SD decrease in IQ); 95% confidence interval: 1.27; 1.10, 1.46) and smoking-related malignancies (1.37; 1.14, 1.64). There was some attenuation following control for mediating variables, particularly smoking and income, but the gradients generally held at conventional levels of statistical significance.
Conclusions: Higher scores on pre-morbid IQ tests are associated with lower risk of later cancer morality. The strength of the relation was partially mediated by established risk factors.