Purpose: We explored correlates of change in cognitive function during a two-year follow-up period among the community-dwelling elderly in Japan, using a population-based prospective approach.
Method: The participants analyzed in the present study were 260 men and 222 women aged 70 to 84 years at baseline, living in an urban Japanese community. Data such as change in cognitive function during two years (calculated by subtracting baseline Mini-Mental State Examination [MMSE] score from follow-up MMSE score: a negative value means a decrease in MMSE scores during the two-year period) as an outcome variable, age, education, hearing and vision problems, IADL deficit (measured by the Tokyo Metropolitan Institute of Gerontology of Index of Competence), problems related to memory complaint, living alone, hemoglobin level, as explanatory variables, and the baseline MMSE score, depressive status (measured by the Mini-International Neuropsychiatric Interview), chronic conditions (hypertension, stroke, and diabetes mellitus) as covariates, were collected during a comprehensive health examination survey for the elderly.
Results and conclusion: We conducted multivariate regression analysis by genders to explore correlates of change in cognitive function. The results showed that higher age (beta = -0.18), presence of hearing problem (beta = -0.21), presence of IADL deficit (beta = -0.15), and memory complaint (beta = -0.20) in men, and higher age (beta = -0.27), low education level (beta = -0.25) and lower hemoglobin level (beta = 0.16) in women, were significantly associated with change in cognitive function when adjusting for the potential confounders. These factors may be reliable predictors for cognitive decline.