[Mini mental state examination and the diagnosis of dementia in Brazil].

The diagnosis of dementia is based on the presence of memory deficits and decline of other cognitive functions. Many scales have been designed to aid the clinician in the assessment of at risk subjects, such as the elderly. The Mini-Mental State Examination (MMSE) is the most widely used of such scales, although its use as an aid to the diagnosis of dementia has not as yet been studied in a Brazilian sample of patients. The current study was designed with 2 main aims: (1) to determine the best cut-off point of the MMSE for the diagnosis of dementia in a sample of elderly subjects assessed in a mental health outpatient unit and; (2) evaluate the impact of age and schooling on MMSE scores. Two hundred and eleven subjects aged 60 or over assessed at the Mental Health Outpatient Unit for the elderly at "Santa Casa de São Paulo" between February 1997 and February 1998 were included in the study. They were assessed with the SRQ-20, MMSE, and a clinical interview for the diagnosis of dementia according to the ICD-10. Seventy patients received the diagnosis of dementia. The MMSE cut-off point of 23/24 (cases/non-cases) was associated with a sensitivity of 84.3% and specificity of 60.3%. MMSE scores were associated with age (r = -0.41, p < 0.001) and schooling (F = 12.69, p < 0.001). Analysis of covariance taking age into account showed that MMSE scores were significantly lower among those with no formal education (F = 10.51, p < 0.001). A cut-off point of 19/20 on the MMSE was associated with sensitivity of 80.0% and specificity of 70.9% for the diagnosis of dementia in this particular group. Subjects with previous school history were better classified to the diagnosis of dementia with the cut-off point of 23/24: sensitivity of 77.8% and specificity of 75.4%. The evaluation of elderly subjects with the MMSE should take education into account, and different cut-off points should be used accordingly as a guideline to the diagnosis of dementia.