Age, education and dementia related deaths. The Norwegian Counties Study and The Cohort of Norway.

An inverse relationship between educational level and dementia has been reported in several studies. In this study we investigated the relationship between educational level and dementia related deaths for cohorts of people all born during 1915-39. The cohorts were followed up from adulthood or old age, taking into account possible confounders and mediating paths. Our study population comprised participants in Norwegian health examination studies in the period 1974-2002; The Counties Study and Cohort of Norway (CONOR). Dementia related deaths were defined as deaths with a dementia diagnosis on the death certificate and linked using the Cause of Death Registry to year 2012. The study included 90,843 participants, 2.06 million person years and 2440 dementia related deaths. Cox regression was used to assess the association between education and dementia related deaths. Both high and middle educational levels were associated with lower dementia related death risk compared to those with low education when follow-up started in adulthood (35-49 years, high versus low education: HR=0.68, 95% confidence interval (CI) 0.50-0.93; 50-69 years, high versus low education: HR=0.52, 95% CI 0.34-0.80). However, when follow-up started at old age (70-80 years) there was no significant association between education and dementia related death. Restricting the study population to those born during a five-year period 1925-29 (the birth cohort overlapping all three age groups), gave similar main findings. The protective effects found for both high and middle educational level compared to low education were robust to adjustment for cardiovascular health and life style factors, suggesting education to be a protective factor for dementia related death. Both high and middle educational levels were associated with decreased dementia related death risk compared with low educational level when follow-up started in adulthood, but no association was observed when follow-up started at old age.