Incidence and risk factors of vascular dementia and Alzheimer's disease in a defined elderly Japanese population: The Hisayama Study

Abstract

Article abstract—We followed 828 nondemented residents of Hisayama Town, Kyushu, Japan, aged 65 years or older (88.3% of the elderly population) for 7 years starting in 1985 in order to determine the type-specific incidence of dementia and its risk factors in the general Japanese population. Only two subjects were lost to the follow-up, during which period 103 subjects developed dementia. Morphologic examination of the brains of 89 subjects (86.4%) was made by autopsy or CT. We made the initial diagnosis of dementia based on the DSM-111-R criteria, with the diagnoses of vascular dementia (VD) being based on the NINDS-AIREN criteria and Alzheimer9s disease (AD) on the NINCDS-ADRDA criteria. The incidence of VD and AD increased with age for both sexes. The age-adjusted total incidence (per 1,000 person-years) of dementia was 19.3 for men and 20.9 for women. The corresponding rates for VD were 12.2 for men and 9.0 for women, and for AD, 5.1 for men and 10.9 for women. Among the VD subjects whose brain morphology we examined, the most frequent type of stroke was multiple lacunar infarcts (42%), but half these subjects lacked a stroke episode in their histories. Multivariate analysis showed that age, prior stroke episodes, systolic blood pressure, and alcohol consumption were significant independent risk factors for the occurrence of VD. In contrast, age and a low score on Hasegawa9s dementia scale were significant risk factors for AD, and physical activity was a significant preventive factor for AD. Our findings suggest that asymptomatic stroke is an important factor in the development of VD, with age, prior stroke episodes, systolic blood pressure, and alcohol consumption being independent risk factors for its occurrence. Age and a low scare on Hasegawa9s dementia scale are significant risk factors for AD, with moderate physical activity having a statistically significant preventive effect.