Objective: To estimate the impact family factors shared by siblings has on the association between length of education and cause-specific mortality in adulthood.
Methods: The study population (n = 871 367) was all Norwegians born 1940-59 having one or more sibling within the cohort and alive in 1990. Length of education was obtained in 1990. Follow-up of deaths was from 1991 to 2008 when participants were aged 32-68 years.
Results: Sixty-five per cent of participants had one or more siblings who had completed a different number of years of formal education. A one-category difference in education was associated with a 30% increase in the hazard rate of death by all causes among men in the cohort analysis and 23% in within siblings analysis, and in women, increases were 22% and 14%, respectively. For cardiovascular disease, increases were 36% and 25% in men and 51% and 36% in women. For lung cancer, they were 48% and 29% in men and 53% and 22% in women. External causes and alcohol-related causes in men were generally similar in both analyses.
Conclusions: This study suggests that at least some of the educational inequalities in all-cause, cardiovascular disease and lung cancer, external and alcohol-related mortality are explained by factors shared by siblings.