The global prevalence of dementia: a systematic review and metaanalysis.
INTRODUCTION Folstein's Mini Mental State Examination (MMSE) is widely used as screening test for cognitive impairment. OBJECTIVE To test a Spanish version of the MMSE in a population of high illiteracy rate. MATERIAL AND METHODS Population-based survey of a stratified random sample of urban and rural residents of five regions of Colombia, followed by neurological and neuropsychological evaluation of suspect cases (phase 2). Dementia was diagnosed using DSM-IV criteria. RESULTS 1,611 subjects age 50 or older filled out both the WHO Protocol for Epidemiologic Studies of Neurological Disorders and a Spanish version of the MMSE; 55.2% of them had three or less years of schooling; 536 individuals with scores below cutoff points were sent to phase 2. Of the population with satisfactory scores in MMSE 366 (34.0%) were evaluated by neurologists to exclude other neurological conditions. Twelve cases of dementia were diagnosed among individuals with scores below cutoff point and one among subjects with high scores. Age-adjusted prevalence was 8.1 per thousand subjects age 50 or over (95% CI: 3.7-12.5); and 34.2 per thousand for ages 75 or over (95% CI: 12.2-56.2). Sensitivity and specificity were 92.3 and 53.7%; 16 of the 19 questions show significant differences (p < 0.001) according to educational level. A gender gap is significant in low educational levels (p < 0.001) but not in subjects with more than five years of schooling. CONCLUSIONS MMSE scores correlated closely with level of education. Low specificity leads to many non-demented subjects with low educational status requiring further investigation.