Introdução: A demência é um dos principais problemas de saúde devido ao rápido crescimento populacional de idosos, com prevalência estimada de 8,5% em estudos latino-americanos. Comprometimento cognitivo leve e Comprometimento Cognitivo sem Demência (CCSD) são designações utilizadas para a condição clínica entre o envelhecimento normal e demência com prevalência ainda insuficientemente determinada em estudos populacionais, particularmente no Brasil. Objetivos: Estabelecer a prevalência de CCSD e de demência em idosos que vivem na comunidade no município de Tremembé, estado de São Paulo, e avaliar a influência de variáveis sociodemográficas. Verificar se a prevalência de demência foi mais elevada entre sexagenários do que em países desenvolvidos. Secundariamente, verificar a adequação dos instrumentos: Questionário de Mudança Cognitiva (QMC), Addenbrooke's Cognitive Examination-Revised (ACE-R) e Montreal Cognitive Assessment (MoCA) para estudos populacionais. Métodos: O estudo populacional foi realizado em fase única com os idosos com 60 anos ou mais residentes no município de Tremembé. Foram selecionados, de forma aleatória, 20% dos domicílios dos idosos de cada setor censitário do município. Além da avaliação clínica e neurológica, questionários foram aplicados aos informantes para avaliar atividade funcional (Informant Questionnaire on Cognitive Decline in the Elderly [IQCODE] e o Questionário de Atividades Funcionais de Pfeffer), e testes neuropsicológicos foram aplicados aos participantes (Bateria Breve do Rastreio Cognitivo, Mini Exame do Estado Mental, teste de fluência verbal e teste do desenho do relógio) juntamente com escalas de avaliação psiquiátrica. Os diagnósticos de CCSD e de demência foram estabelecidos em reunião de consenso entre dois especialistas em neurologia cognitiva, com base no conjunto dos dados obtidos.

Introduction: Dementia is one of the major health issues due to the rapidly growing elderly population, with an estimated prevalence of 8.5% in Latin American studies. Mild cognitive impairment and Cognitive impairment no dementia (CIND) are designations for a clinical condition that lies between normal aging and dementia, and its prevalence is insufficiently determined in population studies, particularly in Brazil. Objectives: To determine the prevalences of CIND and dementia in the communitydwelling elderly living in the municipality of Tremembé, São Paulo state and to evaluate the influence of socio-demographic variables. To verify whether the prevalence of dementia was higher among sexagenarians than in developed countries. Secondarily, to investigate the suitability of the use of Questionnaire of Cognitive Change (QCC), Addenbrooke's Cognitive Examination-Revised (ACE-R) and Montreal Cognitive Assessment (MoCA) in epidemiological studies. Methods: The population study was conducted in a single phase with the elderly aged 60 years or more living in municipality of Tremembé. Twenty percent of the households with elderly were randomly selected from each city's census tract. In addition to clinical and neurological evaluations, questionnaires were applied to informants to assess functional activities (Informant Questionnaire on Cognitive Decline in the Elderly [IQCODE] and Pfeffer Functional Activities Questionnaire); and neuropsychological tests were administered to participants (Brief Cognitive Screening Battery, Mini-Mental State Examination, verbal fluency test and clock drawing test) along with psychiatric rating scales. The diagnoses of dementia and CIND were established in consensus meetings by two neurologists specialized in cognitive neurology, based on the available set of data. Results: Six hundred and thirty subjects (63% women) were examined with mean age of 71.28 (±7.99) and mean years of schooling of 4.90 (±4.54). The prevalence of dementia was 17.5%, higher than in developed countries in all age groups, but the prevalence was even higher among individuals aged between 60 and 69 years. The prevalence of CIND was 19.5%. These prevalences were influenced by age (p< 0.001) and by educational level (p < 0.001). There was no significant gender difference among diagnostic groups (p= 0.166). In multiple comparisons analysis, age was significantly different among the three groups, being higher in the dementia group, lower in the cognitively normal group and intermediate in the CIND group; and education was higher in the normal group compared to the mean years of schooling in the other two groups. Diagnosis of dementia was significantly associated with low socioeconomic status, stroke, previous psychiatric disorder, alcoholism, and epilepsy, while diagnosis of CIND was only associated with stroke. Conclusions: The prevalence of dementia in this study was higher than those reported by other studies, particularly among sexagenarians elderly. The prevalences of dementia and CIND were influenced by age and by educational attainment. The QCC was a suitable instrument for the diagnosis of dementia but the ACE-R and MoCA did not show good accuracies in this population with heterogeneous educational background