OBJECTIVES: This study aims to assess the association of household's and children's education on the risk of type 2 diabetes (T2D) and subsequent death.

STUDY DESIGN: Danish register-based cohort study.

METHODS: In total, 1,021,557 adults were included at their 65th birthday between 2000 and 2018. A multistate survival model was performed to estimate the association of household's and children's education on the transition between the three states: 1) 65th birthday; 2) diagnosis of T2D; and 3) all-cause death.

RESULTS: The incidence rates per 1000 person-years were 9.1 for T2D, 18.4 for death without T2D, and 45.0 for death with T2D. Compared to long household's education and children's education, long household's education combined with either short-medium children's education or no children were associated with a 1.49- (95% confidence interval [CI]: 1.44; 1.54] and 1.69-times (95% CI: 1.61;1.78) higher hazard of T2D, respectively. Short-medium household's education combined with either long children's education or no children were associated with 0.64- (95% CI: 0.62; 0.66) and 0.77-times (95% CI: 0.74; 0.79) lower hazard of T2D, respectively. Compared to long household's education and children's education, any other combination of household's and children's education was associated with higher hazards of death both without and with T2D.

CONCLUSION: Older adults living in households with long education with no children or children with short-medium education had higher hazards of T2D. Households with short-medium education and no children or children with long education were associated with lower hazards of T2D. Both household's and children's education were associated with higher hazard of death without and with T2D.