A longitudinal population study of the mini-mental state examination in the very old: relation to dementia and education.

We examined the Mini-Mental State Examination (MMSE) in relation to the prevalence and incidence of dementia in a representative population sample (n = 494), followed from age 85 to 88. The study included a neuropsychiatric examination (including the MMSE) and a key informant interview conducted by a psychiatrist. Dementia was defined according to the DSM-III-R. Among non-demented 85-year-olds, 75% of those who scored below the cut-off score of 23/24 and 37% of those scoring 24-25 became demented during the follow-up. Non-demented 85-year-olds who did not develop dementia during the following 3 years had a mean decline in MMSE score of 0.6 per year, while those who became demented decreased by 2.3 points per year. A decrease of 4 or more points during the 3 years had a sensitivity of 83% and a specificity of 80% for a diagnosis of dementia. High and low MMSE score at base-line was related to a smaller decrease in MMSE score between age 85 and 88. Higher education was related to higher scores on MMSE at age 85 and to a smaller decline in scores between age 85 and 88. Change in MMSE score in non-demented individuals is small especially among individuals with high education. Non-demented individuals scoring below 26 should receive close attention, as a high proportion are at risk of developing dementia. Copyright 2000 S. Karger AG, Basel