BACKGROUND: Recent studies suggest that trends in cardiovascular risk may result in a decrease in age-specific prevalence of dementia. Studies in rural areas are rare.

OBJECTIVES: To study cohort effects in dementia prevalence and survival of people with dementia in a Swedish rural area.

METHODS: Participants were from the 1995-1998 Nordanstig Project (NP) (n = 303) and the 2001-2003 Swedish National study on Aging and Care in Nordanstig (SNAC-N) (n = 384). Overall 6-year dementia prevalence and mortality in NP and SNAC-N were compared for people 78 years and older. Logistic regression analyses were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) for dementia occurrence using the NP study population as the reference group. Cox regression models were used to analyze time to death.

RESULTS: The crude prevalence of dementia was 21.8% in NP and 17.4% in SNAC-N. When the NP cohort was used as the reference group, the age- and gender-adjusted OR of dementia was 0.71 (95% CI 0.48-1.04) in SNAC-N; the OR was 0.47 (0.24-0.90) for men and 0.88 (0.54-1.44) for women. In the extended model, the OR of dementia was significantly lower in SNAC-N than in the NP cohort as a whole (0.63; 0.39-0.99) and in men (0.34; 0.15-0.79), but not in women (0.81; 0.46-1.44). The Cox regression models indicated that the hazard ratio of dying was lower in the SNAC-N than NP population.

CONCLUSIONS: Trends toward a lower prevalence of dementia in high-income countries seem to be evident in this Swedish rural area, at least in men.