Reliable change indices for the 37-item version of the MMSE in Spanish older adults

Abstract Objective The aim of this study was to establish reliable change and regression-based change score norms on the 37-item version of MMSE in older Spanish adults at the three-year follow-up. Method All subjects of this research were selected from the Neurological Disorders in Central Spain (NEDICES), a prospective population-based cohort study of older adults (65 years and over). Of the 4208 individuals free from neurological disorders in this cohort, 2073 participants completed the MMSE-37 at baseline and at the three-year follow-up. Reliable Change Indices were computed for the 80, 90% and 95 confidence intervals (CIs). Multiple regression analyses were used to derive equations for predicting MMSE-37 post-test scores taking into account baseline scores, time to follow-up and sociodemographic factors. Results The MMSE-37 obtained a marginal test-retest reliability (.70). The results showed significant effects of education, age, and sex on the MMSE-37 change scores. After correcting for regression to the mean, at least a 6-point change on MMSE-37 (three-years follow-up) is required to be classified as reliable (90% CI). Conclusions These findings demonstrate that the MMSE-37 is a reliable test-retest measure whose change scores are significantly influenced by sociodemographic factors. Importantly, small changes on this measure require a cautious interpretation.